Market review: Pulse oximeters in primary and prehospital care

Transcription

Market review
Pulse oximeters in primary and
prehospital care
CEP10066
March 2010
Contents
2
Introduction ............................................................................................. 3 Key device considerations....................................................................... 6 Fingertip pulse oximeters ........................................................................ 7 Handheld pulse oximeters ..................................................................... 16 Wrist-worn pulse oximeters ................................................................... 26 Bedside / desktop pulse oximeters ........................................................ 28 Choosing a pulse oximeter .................................................................... 34 Acknowledgements ............................................................................... 37 References............................................................................................ 40 Appendix 1: Supplier contact details ..................................................... 45 Appendix 2: Summary of clinical trials ................................................... 49 Author and report information................................................................ 91 CEP10066: March 2010
Introduction
3
CEP market reviews provide comparative product specifications for healthcare
products available to the UK market within a defined category. They supplement CEP
buyers’ guides, which offer more general advice on the technical, operational, and
economic considerations to be taken into account when selecting an appropriate
product.
This market review is intended to assist in the selection of pulse oximeters for use
within primary and prehospital care. However, some or all of the devices may also be
of use in secondary care.
Scope
This market review provides information on pulse oximeters that can be used across
healthcare services, particularly within primary and prehospital care. The 85 pulse
oximeter models in this review are representative of the UK market at the beginning
of 2010. They were grouped into four categories:
• Fingertip: a pulse oximeter with an integrated fingertip probe which incorporates a
display and records results (list price range £50.00 - £415.00).
• Handheld: a pulse oximeter that displays and records results and is intended to
be held in the hand during normal use. Probes are connected to the unit via a
cable (list price range £143.10 - £5800.00).
• Wrist-worn: a pulse oximeter that displays and records results and is intended to
be worn on the wrist during normal use. Probes are connected to the unit via a
cable (list price range £558.57 - £695.00).
• Bedside/desktop: a pulse oximeter that sits on a desk or is supported by a stand
during normal use. Probes are connected to the unit via a cable or wireless link
(list price range £785.00 - £2703.75).
No multi-parameter monitors (ie vital sign monitors and patient monitors) are included
in this market review, as these systems are predominantly used in secondary care.
Hospital grade non-invasive blood pressure monitors with pulse oximetry were
included in a previous buyers’ guide [1]. Fetal oximeters, cerebral oximeters,
oximetry recorders and oximeters intended for non-NHS markets (eg sports, aviation
and consumer-use) were also excluded from this market review.
This market review is concerned with the monitoring of oxygen saturation noninvasively. Direct arterial oxygen saturation is measured invasively using co-oximetry,
covered in previous CEP reports [2, 3].
This market review should be read in conjunction with the associated buyers’ guide
[4].
CEP10066: March 2010
Introduction
4
The clinical and cost effectiveness of pulse oximeters in primary and prehospital care
have been previously reviewed by CEP [5].
Method
Product identification
During 2009/10 we carried out a review of pulse oximeters available on the UK
market that could be used in primary and prehospital care, drawing on information
from medical literature, marketing literature from known suppliers and internet
searches. We identified 85 models from 21 manufacturers. These comprised:
• 28 fingertip pulse oximeters
• 34 handheld pulse oximeters
• 3 wrist-worn pulse oximeters
• 20 bedside/desktop pulse oximeters
Collation of product information
Product information was derived from the manufacturers’ and suppliers’ product
brochures or from specifications published in user manuals. We have presented this
information (tables 2, 4, 6 and 8) in a common, consistent format to allow easy
comparison between devices. Devices are ordered alphabetically by manufacturer
and by model. Where the manufacturer or supplier did not provide data not specified
is shown.
Evidence of accuracy
The performance of pulse oximeters is defined in the standard BS EN ISO 9919:2009
[6], which specifies that accuracy should be supported by a clinical trial. The standard
originated in 1992 [7], but no specification for SpO2 accuracy was defined. However,
it did state that manufacturers must disclose accuracy and test methods upon
request. Specifications for SpO2 accuracy were introduced in the 2005 revision [8].
We searched for peer-reviewed literature relating to the performance of the pulse
oximeters included in the market review. The following questions were considered:
• Has the accuracy of the pulse oximeter been determined as specified in BS EN
ISO 9919:2009 (or an earlier version)?
• Has the performance of the pulse oximeter been compared with other pulse
oximeters and/or the gold standard analysis of arterial blood using co-oximetry?
A literature search on MEDLINE was completed in January 2010 using pulse
oximeter makes, models and technology as search terms. Relevant literature was
also requested from manufacturers and suppliers.
Introduction
5
Twenty-nine peer-reviewed articles were identified: 13 considered the accuracy of
pulse oximeters [9-21] and 16 considered the performance of pulse oximeters [2237]. Accuracy testing is where SpO2 measurements made by the pulse oximeter
under test are directly compared with SaO2 measurement made using a co-oximeter.
Performance testing is where pulse oximeters are compared against each other.
In addition to peer-reviewed articles, manufacturers / suppliers were asked to provide
any unpublished or non-peer reviewed evidence (including meeting abstracts and
conference papers) on accuracy of pulse oximeters. Nine unpublished reports [38-46]
and nine non-peer reviewed abstracts [47-55] were received.
The evidence relating to accuracy and performance of pulse oximeters has been
summarised in tables 11-17, appendix 2. Where models are claimed to use the same
pulse oximetry (SpO2) technology as a device for which accuracy and/or performance
has been demonstrated in the literature, equivalent performance has been assumed.
Key device considerations
6
Table 1 contains device characteristics to be considered when purchasing pulse
oximeters. This table should be used in conjunction with the comparative product
information tables (tables 2, 4, 6 and 8).
Table 1. Device considerations
Feature
Description
SpO2 technology
Manufacturers have developed different algorithms to determine SpO2
from the measured absorbance of red and infrared light. Algorithms use
different and usually proprietary methods to reduce the effect of low
amplitude and artefact, such as movement and ambient light, on
measurement accuracy.
Sensor wavelengths
Red and infrared (or near infrared) light wavelengths used to determine
SpO2.
SpO2 range and accuracy
The range of oxygen saturation for which the pulse oximeter measures
within the stated accuracy.
BS EN ISO 9919:2009 [6] states that SpO2 accuracy should be less
than ±4% over the range of 70% to 100%. Accuracy should be
supported by a clinical study where SpO2 measurements have been
compared with SaO2 measurements made using a co-oximeter.
Pulse rate range and
accuracy
The range of pulse rate or heart rate for which the pulse oximeter
measures within the stated accuracy.
BS EN ISO 9919:2009 [6] states that pulse rate accuracy should be
supported by evidence of comparison with a reference method of
measuring heart rate, eg electronic pulse simulator or ECG heart rate.
Display type
This can be an integrated LED or LCD display. Some displays are
monochrome whilst others are colour. Purchasers should consider
whether high visibility or brightly coloured displays are important for their
application.
Power supply
Devices in this review are powered by mains or battery, some of which
are rechargeable by connecting to a mains adaptor or via USB.
Battery capacity
Typical length of continuous operation with new or fully charged battery.
Other features
Other features such as display of plethysmograph waveform, memory
options, software, automatic power off etc.
Size and weight
These factors are important when considering the environment the
device is to be used in, eg in GP surgery, by paramedic in the field etc.
List price
We have shown the list price excluding VAT as a guide only; price paid
will depend on various commercial factors.
Accessories and prices
Accessories include sensors, mains adaptor or charger, printer options,
carry case etc. We have shown the current accessory list prices
excluding VAT. For some manufacturers/suppliers, prices vary
depending on quantity.
Most oximeters have a wide range of sensors which could not be listed
due to space limitations. Users should contact suppliers to establish the
full list of accessories available.
Fingertip pulse oximeters
7
Table 2. Fingertip pulse oximeters
Beijing Choice
MD300-C
Beijing Choice
MD300-C5
Beijing Choice
MD300-D
Beijing Choice
M-Pulse
Beijing Choice
Beijing Choice
Beijing Choice
Beijing Choice
not specified
not specified
not specified
not specified
99%-70%
99%-0%
99%-75%: ±2%;
74%-50%: ±3%;
<50% unspecified
99%-70%
±2%
99%-80%: ±2%
80%-70%: ±3%
<70% unspecified
99%-80%: ±2%
80%-70%: ±3%
<70% unspecified
30-240 bpm
30-235 bpm
30-235 bpm
0-254 bpm
±2 bpm
±2 bpm
±2 bpm
±2 bpm or ±2%
Display type
LED
dual colour OLED
dual colour OLED
dual colour OLED
Power supply
2 x AAA alkaline
batteries
2 x AAA alkaline
batteries
2 x AAA alkaline
batteries
2 x AAA alkaline
batteries
30 hours
30 hours
continuous use
30 hours
continuous use
approx. 30 hours
8 second auto power
off
rotatable display;
plethysmogram
waveform display;
8 second auto power
off; 10 level adjustable
brightness
4 way rotatable display;
plethysmogram
waveform display;
8 second auto power
off; 10 level adjustable
brightness
finger size range:
8-26mm;
plethysmogram
waveform display;
10 level adjustable
brightness
58 x 34 x 32 mm
50 x 28 x 28 mm
58 x 34 x 32 mm
58 x 32 x 34 mm
0.080 kg
(inc. battery)
0.042 kg
0.090 kg
0.050 kg (inc batteries)
£50.00
£90.00
£80.00
£77.50
SpO2 technology
Sensor wavelengths
100%-0%
SpO2 range &
accuracy
Pulse rate range &
accuracy
Battery capacity
Other features
Size (H x W x D)
Weight
List price (exc. VAT)
Accessories &
prices
carry case £4.95
8
Table 2. Fingertip pulse oximeters (continued)
SpO2 technology
Sensor wavelengths
SpO2 range &
accuracy
Beijing Choice
M-Pulse Lite
Comdek
MD-650P
Traveler II
Daray
V407
Daray
V408
Beijing Choice
Comdek
Daray
Daray
not specified
not specified
not specified
not specified
99%-70%
100%-0%
99%-80%: ±2%
80%-70%: ±3%
100%-80%: ±2%
<80% unspecified
99%-35%
range not specified
±1%
100%-70%: ±2%
<70% unspecified
30-235 bpm
30-250 bpm
range not specified
range not specified
±2 bpm or ±2%
±1% of full scale
±1 bpm
±2 bpm
Display type
LED
LED
OLED
LED
Power supply
2 x AAA alkaline
batteries
3 x AAA batteries
2 x AAA alkaline or
rechargeable batteries
2 x AAA alkaline or
not specified
not specified
not specified
approx. 100 hours
10 seconds
auto power off
adjustable brightness;
auto power off;
plethysmogram
waveform display
8 second
auto power off
58 x 32 x 34 mm
73 x 45 x 43 mm
not specified
not specified
0.060 kg
not specified
<0.050 kg
£65.00
£123.20
£149.00
£69.00
Pulse rate range &
accuracy
Battery capacity
Other features
Size (H x W x D)
Weight
Accessories &
prices
carry case £4.95
9
Daray
V409
Daray
V501
Edan
H10
Konica Minolta
Pulsox-1
Daray
Daray
Edan
Konica Minolta
not specified
not specified
not specified
red 659-677 nm
infrared 900-990 nm
99%-35%
±2%
99%-80%: ±2%
79%-70%: ±3%
<70% unspecified
range not specified
range not specified
30-240 bpm
±2 bpm
±2 bpm
±3% or ±2 bpm
30-100 bpm: ±2 bpm
101-230 bpm: ±2% of
reading
Display type
OLED
OLED
128 x 64 pixel
dual colour OLED
backlit LCD
with light sensor
Power supply
2 x AAA alkaline or
2 x AAA batteries
2 x AAA alkaline
batteries
1 x AAA alkaline
battery
approx. 100 hours
50 hours
25 hours continuous
approx. 55 hours
rotatable display;
8 second auto power
off; plethysmogram
waveform display
auto rotation display;
8 second auto power
off
rotatable display;
auto power off;
plethysmogram
waveform display;
10 level adjustable
brightness
1 minute display hold;
neck strap included
not specified
not specified
57 x 32 x 31 mm
56 x 35 x 33 mm
<0.050 kg
approx. 0.050 kg
0.060 kg
(inc. batteries)
0.049 kg
(inc. battery)
£99.00
£149.00
£58.61
£220.00
SpO2 technology
Sensor wavelengths
SpO2 range &
accuracy
99%-35%
99%-70%: ±2%
<70% unspecified
99%-35%
100%-0%
100%-70%, 1sd: ±2%
<70% unspecified
30-230 bpm
Pulse rate range &
accuracy
Battery capacity
Other features
Size (H x W x D)
Weight
Accessories &
prices
10
SpO2 technology
Sensor wavelengths
SpO2 range &
accuracy
Konica Minolta
Pulsox-2
Mediaid
100 / 100C
Nissei
BO-800
Nonin
9500 Onyx
Konica Minolta
Mediaid
Nissei
Nonin PureSAT®
not specified
red 660 nm
infrared 910 nm
not specified
red 660 nm
infrared 910 nm
100%-0%
100%-70%, 1sd: ±2%
<70% unspecified
100%-20%
100%-70%: ±2%
69%-60%: ±3%
<60% unspecified
100%-0%
accuracy not specified
100%-0%
100%-70%: ±2 digits
<70% unspecified
25-250 bpm
Pulse rate range &
accuracy
20-250 bpm
±2 bpm
25-200 bpm: ±2 bpm
or 2%, whichever is
greater
>200 bpm: ±3%
Display type
backlit LCD
LED
backlit LCD
LED
Power supply
2 x AAA alkaline
batteries
1 x AA alkaline battery
2 x AAA batteries
2 x AAA alkaline
batteries
72 hours
approx. 1200 spot
checks
up to 200 hours
(without backlight);
18,000-24,000
spot checks
up to 18 hours
(continuous);
1600 spot checks
auto power off
finger size range:
8-26mm;
auto power off
Battery capacity
30-240 bpm
18-300 bpm
accuracy not specified
±3% ±1 digit
model 100: integrated
finger sensor;
Other features
Size (H x W x D)
Weight
integrated finger
sensor; neck strap
included
model 100C has an
additional cable adaptor
module & sensor,
11.5 hour, 20 patient
internal memory;
infrared interface &
PC software
69 x 60 x 28 mm
120 x 47 x 25 mm
51 x 30 x 26 mm
57 x 33 x 33 mm
0.057 kg
(exc. batteries)
0.098 kg
0.046 kg
(inc. batteries)
0.060 kg
(inc. batteries)
£311.95
100: £215.00
100C: £415.00
£169.00
£169.00
Accessories &
prices
carrying case
11
SpO2 technology
Sensor wavelengths
Nonin
9550 Onyx II
Nonin
9560 Onyx II
Nonin
GO2
Nonin
Medair
OxyCheck
Nonin PureSAT®
Nonin PureSAT® with
SmartPoint™ technology
Nonin PureSAT®
Nonin PureSAT®
red 660 nm
infrared 910 nm
red 660 nm
infrared 910 nm
red 660 nm
infrared 910 nm
red 660 nm
infrared 910 nm
100%-0%
100%-0%
100%-70%: ±2 digits
<70% unspecified
normal & low perfusion:
100%-70%: ±2 digits
<70% unspecified
normal & low perfusion:
100%-70%: ±2 digits
<70% unspecified
18-321 bpm
18-321 bpm
18-321 bpm
18-321 bpm
20-250 bpm: ±3 digits
20-250 bpm: ±3%
low perfusion:
low perfusion:
low perfusion:
Display type
LED
LED
backlit LCD
LCD
Power supply
2 x AAA alkaline
batteries
2 x AAA alkaline
batteries
1 x AAA alkaline
battery
2 x AAA alkaline
batteries
Battery capacity
approx. 21 hours
(continuous);
2500 spot checks
approx. 600 spot
checks
approx. 21 hours
(continuous);
2400 spot checks
approx. 18 hours
(continuous);
1600 spot checks
Other features
finger size range:
8-26mm;
auto power off
remote patient
monitoring via
Bluetooth® 2.0
wireless technology;
range up to 100m;
minimum 20 measurement internal memory
auto on/off; 3 colours
(blue, orange or green)
finger size range:
8-26mm;
auto on/off;
Size (H x W x D)
56 x 33 x 32 mm
not specified
not specified
not specified
0.054 kg
(inc. batteries)
0.060 kg
(inc. batteries)
not specified
not specified
£229.00
£399.50
£89.50
£69.50
carrying case;
20” lanyard; clip holder
soft carry case £8.50
SpO2 range &
accuracy
Pulse rate range &
accuracy
Weight
Accessories &
prices
100%-0%
carrying case
100%-0%
100%-70%: ±2 digits
<70% unspecified
12
Rossmax
SB-220
SeQual
SmartPulse
Smiths Medical
BCI Digit
SPO Medical
PulseOx 5500
Rossmax
SeQual
BCI
SPO Medical
Reflective Pulse
Oximetry (RPO)
technology
red 660 nm
infrared 880-905 nm
not specified
not specified
not specified
99%-35%
99%-9%
8 pulse averaging
99%-0%
8 beat averaging
99%-70%: ±2%
<70% unspecified
99%-70%: ±2%
<70% unspecified
30-254 bpm
8 pulse averaging
30-254 bpm
8 second averaging
±3 bpm
±2% or ±2 bpm,
whichever is greater
±2% or ±2 bpm,
±3% or ±3 digits
Display type
dual colour OLED
multi-colour LED
LED
backlit LCD
Power supply
2 x AAA alkaline
batteries
2 x AAA alkaline
batteries
2 x AAA alkaline
batteries
1 x ½AA 3.6V lithium
battery
Battery capacity
approx. 16 hours
(continuous)
50 hours
(continuous)
approx. 16 hours
(continuous);
1400 spot checks
approx. 1000 hours or
40 days continuous use
Other features
SpO2 <90% alarm;
auto power off
8 second auto power
off; audio & visual
alerts; detachable
lanyard & carry case
8 second auto power
off
automatic on/off;
includes lanyard
Size (H x W x D)
63.5 x 35 x 34 mm
60 x 56 x 30 mm
57 x 43 x 38 mm
74 x 41 x 40 mm
0.037 kg
0.050 kg
(inc. batteries)
0.085 kg
(inc. batteries)
0.055 kg
(inc. battery)
£84.25
£164.52
£145.00
£115.00
SpO2 technology
Sensor wavelengths
SpO2 range &
accuracy
Pulse rate range &
accuracy
Weight
99%-70%: ±2%
69%-35% unspecified
30-250 bpm
Accessories &
prices
99%-40%
99%-70%: ±2% or
±2 digits
<70% unspecified
40-250 bpm
belt pouch £7.50
13
SpO2 technology
Sensor wavelengths
SpO2 range &
accuracy
SPO Medical
PulseOx 6000
Viamed
VM-2101
Viamed
VM-2105
SPO Medical
Reflective Pulse
Oximetry (RPO) with
AutoSpot™ technology
Viamed
Viamed
not specified
red 660 nm
infrared 905 nm
red 660 nm
infrared 905 nm
99%-40%
100%-0%
100%-0%
99%-70%: ±2% or
±2 digits
<70% unspecified
100%-70%: 1sd: ±2%;
Arms = <3%
<70% unspecified
100%-70%: 1sd: ±2%;
Arms = <3%
<70% unspecified
20-300 bpm
20-300 bpm
±3% or ±3 digits
≤100 bpm: ±1 bpm
>100 bpm: ±1%
≤100 bpm: ±1 bpm
>100 bpm: ±1%
Display type
backlit LCD
128 x 96 pixel
multi-colour OLED
128 x 96 pixel
multi-colour OLED
Power supply
1 x ½AA 3.6V lithium
battery
2 x AAA alkaline
batteries
2 x AAA alkaline
batteries
approx. 500 hours
continuous use
approx. 24 hours
continuous use
approx. 24 hours
continuous use
180° bi-directional
display; automatic
on/off; includes
lanyard & belt pouch
4 way rotatable display;
plethysmogram
waveform display;
includes lanyard
soft silicone housing;
4 way rotatable
display;
plethysmogram
waveform display;
includes lanyard
74 x 41 x 30 mm
57 x 33 x 27 mm
65 x 50 x 34 mm
0.050 kg
(inc. battery)
approx. 0.050 kg
(inc. batteries)
approx. 0.049 kg
(inc. batteries)
£135.00
£120.00
£225.00
Pulse rate range &
accuracy
Battery capacity
Other features
Size (H x W x D)
Weight
40-250 bpm
Accessories &
prices
lanyard £2.00; hard carry case £4.00;
soft carry case £4.00
14
Unpublished report
Non-peer reviewed
article
Peer-reviewed
article (comparison
with other pulse
oximeters)
Peer-reviewed
article (comparison
with co-oximetry)
Table 3. Evidence of clinical study to confirm accuracy of fingertip pulse oximeters
Manufacturer
Model
Beijing Choice
MD300-C
no evidence
Beijing Choice
MD300-C5
no evidence
Beijing Choice
MD300-D
no evidence
Beijing Choice
M-Pulse
no evidence
Beijing Choice
M-Pulse Lite
no evidence
Comdek
MD-650P Traveler II
no evidence
Daray
V407
no evidence
Daray
V408
no evidence
Daray
V409
no evidence
Daray
V501
no evidence
Edan
H10
no evidence
Konica Minolta
Pulsox-1
9
[45]
Konica Minolta
Pulsox-2
9
[45]
Mediaid
100/100C
9
[44]
Nissei
BO-800
Nonin
9500 Onyx
9
[17, 23]
9
[39-41]
Nonin
9550 Onyx II
9
[17, 23]
9
[39-41]
Nonin
9560 Onyx II
9
[17, 23]
9
[39-41]
Nonin
GO2
9
[17, 23]
9
[39-41]
Nonin
Medair OxyCheck
9
[17, 23]
9
[39-41]
Rossmax
SB-220
no evidence
SeQual
SmartPulse
no evidence
Smiths Medical
BCI Digit
SPO Medical
PulseOx 5500
9
[41, 43]
SPO Medical
PulseOx 6000
9
[41, 43]
Comment
no evidence
9
[37]
15
Unpublished report
Non-peer reviewed
article
Peer-reviewed
article (comparison
with other pulse
oximeters)
Peer-reviewed
article (comparison
with co-oximetry)
Table 3. Evidence of clinical study to confirm accuracy of fingertip pulse oximeters (continued)
Manufacturer
Model
Viamed
VM-2101
no evidence
Viamed
VM-2105
no evidence
Comment
Handheld pulse oximeters
16
Table 4. Handheld pulse oximeters
SpO2 technology
Sensor wavelengths
Beijing Choice
MD300K1-E
Bionet
Oxy9
Bitmos
Medizintechnik
SAT 800
Bitmos
Medizintechnik
SAT 801
Beijing Choice
Bionet
Masimo SET®
Masimo SET®
not specified
red 680 nm
infrared 890 nm
red 660 nm
infrared 940 nm
red 660 nm
infrared 940 nm
adults & children no motion:
100%-70%: ±2%
69%-0%: unspecified
100%-70%
SpO2 range &
accuracy
100%-80%: ±2%
79%-70%: ±3%
≤69%: unspecified
100%-80%: ±2%
80%-70%: ±3%
69%-40%: unspecified
neonates - no motion:
100%-70%: ±3%
69%-0%: unspecified
adults & children with motion
100%-70%: ±3%
69%-0%: unspecified
Pulse rate range &
accuracy
Display type
Power supply
Battery capacity
Other features
Size (H x W x D)
Weight
30-235 bpm
30-100 bpm: ±2 bpm
101-235 bpm: ±2%
30-300 bpm
±1 bpm
25-240 bpm
±3 bpm (no motion)
±5 bpm (with motion)
LED
monochrome LCD
backlit monochrome LCD
2 x AA alkaline batteries
3 x AA rechargeable
NiMH batteries with
external mains adaptor
2 x AA batteries
approx. 30 hours
12-20 hours
20 hours (typical)
alarms; 100 patient,
72 hour internal
memory; shock-proof
& highly visible; carry
case & IV pole strap
perfusion index;
integrated rubber
protectors; USB
interface
2 x AA batteries
(battery version);
rechargeable Li-Ion
batteries
(1.6Ah & 3.2Ah
Li-Ion versions);
(optional on battery
version)
20 hours (typical)
(battery version);
15 or 30 hours
(depending on Li-Ion
version)
as for Sat800 with
additional prioritised
alarms & internal
memory with PC
analysis software
130 x 61 x 30 mm
133 x 65 x 30 mm
128 x 85 x 46 mm
not specified
not specified
0.230 kg
£249.00
price not provided
Accessories &
prices
£499.00
£799.00
external mains adaptor; table stand
17
Table 4. Handheld pulse oximeters (continued)
SpO2 technology
Sensor wavelengths
SpO2 range &
accuracy
Pulse rate range &
accuracy
Display type
Comdek
MD-600P /
MD-610P
Comdek
MD-630P
Comdek
MD-670P
Comdek
MD-680P
Comdek
Comdek
Comdek
Comdek
not specified
not specified
not specified
not specified
100%-80%: ±2%
79%-0%: unspecified
100%-80%: ±2%
79%-0%: unspecified
100%-80%: ±2%
79%-65%: ±3%
64%-0%: unspecified
100%-80%: ±1%
79%-0%: unspecified
30-250 bpm
±1% of full scale
30-250 bpm
±1% of full scale
30-250 bpm
±1%
30-250 bpm
±1% of full scale
LED
LED
backlit monochrome
LCD
LED
4 x AA batteries;
4 x AA rechargeable
batteries with external
mains adaptor
4 x AA batteries;
4 x AA rechargeable
mains adaptor
3 x AAA batteries
20 hours
not specified
not specified
not specified
MD-600P only:
alarms
built-in thermal printer;
alarms; 24 hour
internal memory;
USB interface
alarms; 72 hour
internal memory with
trends; USB interface;
PC analysis software
integrated finger
probe; pulse rate
alarm; 1 minute
auto power off
172 x 90 x 36 mm
155 x 88 x 44 mm
170 x 88 x 44 mm
90 x 62 x 22 mm
0.240 kg
0.300 kg
0.250 kg
0.090 kg
MD-600P: £316.80
MD-610P: £299.20
£519.20
£413.60
£158.40
both models:
4 x AA batteries
Power supply
Battery capacity
Other features
Size (H x W x D)
Weight
MD-600P only:
4 x AA rechargeable
mains adaptor
Accessories &
prices
18
Criticare
503DX
miniSPO2T
Daray
V60
Daray
V202
Edan
H100B
Criticare Systems Inc
(CSI) DOX™
Daray
Daray
Edan
not specified
not specified
red 660 nm
infrared 940 nm
not specified
99%-1%
99%-70%: ±2%
69%-40%: ±3%
<40%: unspecified
100%-0%
adult/paed, no motion:
100%-70%: ±2%
neonate, no motion:
100%-70%: ±3%
69%-0%: unspecified
99%-35%
99%-70%: ±2%
<69%: unspecified
100%-0%
adult: 100%-70%:
±2 digits
neonate: 100%-70%:
±3 digits
<70%: unspecified
20-300 bpm
±1 bpm or ±1%,
25-250 bpm
no motion: ±3 bpm
30-250 bpm
±2 bpm
30-254 bpm
±3 digits
Display type
LED
160 x 96 pixel
1.8” colour TFT LCD
LED
backlit monochrome
LCD
Power supply
4 x AA alkaline
batteries
2 x AA batteries
3 x AA alkaline or
4 x AA alkaline
batteries or
rechargeable NiMH
battery pack
24 hours
not specified
7 hours
48 hours (alkaline) or
36 hours
(rechargeable)
alarms; display
direction adjust;
24 hour internal
memory; data transfer
to PC software
3 minute auto power
off; alarms
alarms; 300 hour
internal memory; realtime clock display;
graphical & tabular
trends; auto power off;
battery charger stand;
serial interface for data
transfer
146 x 91 x 33 mm
92 x 82 x 22 mm
120 x 63 x 32 mm
160 x 70 x 38 mm
Weight
0.280 kg
0.130 kg
0.165 kg
£495.00
price not provided
£269.00
£143.10
adult sensor £85.00
paediatric sensor
£85.00
neonatal sensor £85.00
ear clip sensor £85.00
additional batteries &
charger £29.95
reusable paediatric
finger sensor £52.00
SpO2 technology
Sensor wavelengths
SpO2 range &
accuracy
Pulse rate range &
accuracy
Battery capacity
Other features
Size (H x W x D)
Accessories &
prices
19
SpO2 technology
Sensor wavelengths
SpO2 range &
accuracy
GE Healthcare
TuffSat
Huntleigh
smartsigns
MiniPulse
Masimo
Rad-5 /
Rad-5v
Masimo
Rad-57
GE TruTrak®
BCI
Masimo SET®
Masimo SET®
upgradeable to
Masimo Rainbow® SET
red 650-670 nm
infrared 930-950 nm
not specified
red 660 nm
infrared 940 nm
red 660 nm
infrared 940 nm
100%-1%
Rad-5: 2, 4, 8, 10, 12,
14 or 16 sec averaging
Rad-5v: 8 sec averaging
100%-1%
2, 4, 8, 10, 12, 14 or
16 sec averaging
100%-0%
12 sec averaging
99%-0%
8 sec averaging
100%-70%: ±2 digits
<70%: unspecified
99%-70%: ±2%
<70%: unspecified
motion or low perfusion:
±2 digits (adult/paed)
±3 digits (neonate)
25-240 bpm
Pulse rate range &
accuracy
40-255 bpm
40-100bpm: ±2 bpm
100-255bpm: ±2%
30-254 bpm
±2 bpm or ±2%
Display type
backlit monochrome
LCD
Power supply
Battery capacity
100%-70%:
no motion: ±3 digits
(adult/paed/neonate)
100%-70%: no motion
or low perfusion:
motion: ±3 digits
(adult/paed/neonate)
25-240 bpm
no motion & low
perfusion: ±3 digits
no motion & low
motion: ±5 digits
motion: ±5 digits
LED
LED
LED
4 x AA alkaline
batteries
4 x AA alkaline batteries
(model MP1)
or rechargeable NiMH
cells & desktop charger
(model MP1R)
4 x AA alkaline
batteries
4 x AA alkaline
batteries
17-20 hours
>120 hours
>30 hours
>8 hours
perfusion index; alarms
additional Rainbow®
Rad-5 only: FastSat®
SET parameters:
rapid change & Max
57c carboxyindication; 72 hour
haemoglobin (%SpCO),
internal memory with PC 57m methaemoglobin
output; sleep mode;
(%SpMet) & perfusion
power-on user
index; alarms; 72 hour
configuration
internal memory
pulse strength index;
5 minute auto power
off; 64 minute internal
memory; infrared
printing; protective
rubber bumper
alarms; auto power off;
integrated probe
storage
150 x 70 x 30 mm
140 x 75 x 25 mm
158 x 76 x 36 mm
158 x 76 x 36 mm
Weight
0.257 kg
0.300 kg
0.320 kg
0.370 kg
£555.00
MP1: £423.00
MP1R: £482.00
Rad-5 £550.00
Rad-5v £450.00
Rad-57c £2995.00
Rad-57m £2995.00
Rad-57cm £5800.00
infrared printer
protective cover;
IV pole attachment;
carry case; desk stand
reusable sensor range;
protective boot with
stand
7 colour-choice
protective rubber boot
with stand
Other features
Size (H x W x D)
Accessories &
prices
20
SpO2 technology
Sensor wavelengths
Mediaid
M34
Mediaid
M5340
Mindray
PM-50
Mindray
PM-60
Mediaid
Mediaid
Mindray
Mindray
red 660±2 nm
infrared 910±10 nm
red 660 nm
infrared 910 nm
not specified
not specified
100%-0%
100%-0%
7, 9 & 11 sec averaging
(sensitivity: High,
Med & Low)
100%-0%
4 beat averaging
100%-0%
100%-70%: ±2%
69%-60%: ±3%
<60%: unspecified
100%-70%:
±2% (adult/paed)
±3% (neonate)
69%-0%: unspecified
25-250 bpm
±2 bpm or ±2%,
32-250 bpm
±2 bpm
25-254 bpm
±2 bpm
18-300 bpm
±3 bpm (no motion)
±5 bpm (motion)
Display type
LED
LED
blue monochrome
LCD
2.4” colour TFT LCD
Power supply
9V rechargeable
lithium ion battery
rechargeable NiCad
battery with external
mains adaptor
4 x AA alkaline or
3 x AA batteries or
rechargeable Li-ion
charger stand
21 hours
(in sleep mode)
12 hours
approx. 15 hours
36 hours (standard)
or 24 hours
(rechargeable)
alarms; intermittent,
automatic & sleep
recording modes; up to
136 hour internal
memory; data transfer
to PC via USB or
printer via infrared
interfaces
alarms; data transfer
via serial interface
100 patient, 200
measurement internal
memory; auto power off;
serial RS-232 data
transfer
alarms; spot check &
continuous modes;
96 hour, 99 patient,
4000 measurement
internal memory;
infrared data transfer
140 x 76 x 27 mm
197 x 105 x 38 mm
140 x 65 x 32 mm
120 x 55 x 30 mm
Weight
0.202 kg
0.516 kg
0.130 kg
0.300 kg
£425.00
£395.00
£199.00
£299.00
reusable & disposable
sensors;
infrared printer;
rubber boot £35.00;
IV pole clamp £35.00
sensors;
table/wall mount;
IV pole clamp £35.00
SpO2 range &
accuracy
Pulse rate range &
accuracy
Battery capacity
Other features
Size (H x W x D)
Accessories &
prices
100%-70%: ±2%
<70%: unspecified
100%-70%:
±2% (adult/paed)
±3% (neonate)
69%-0%: unspecified
carrying case; protective
cover; battery charger
kit; pole mounting kit;
data management kit
21
Nellcor
Oximax N-65
SpO2 technology
Sensor wavelengths
Nellcor OxiMax® with
LoSat™
red 660 nm
infrared 900 nm
100%-1%
SpO2 range &
accuracy
100%-70% (±1sd):
adult & low perfusion:
±2 digits
neonates: ±3 digits
Nonin
PalmSAT
2500/2500A
Nonin
8500 series
Smiths Medical
BCI 3301
Nonin PureSAT®
Nonin PureSAT®
BCI
red 660 nm
infrared 910 nm
100%-0%
red 660 nm
infrared 910 nm
not specified
100%-70%:
no motion:
100%-70% (adult):
±2 to ±4 digits
(depending on sensors)
motion or low perfusion:
100%-0%
95%-70% (neonate):
±3 digits
99%-0%
8 beat averaging
99%-70%: ±2%
<70%: unspecified
<70%: unspecified
18-300 bpm
Pulse rate range &
accuracy
20-250 bpm
±3 digits including
low perfusion
no motion:
18-300 bpm ±3 digits
motion:
18-300 bpm
±3% ±1 digit
±2% or ±2 bpm
LED
LED
6 x AA alkaline
batteries
3 x C alkaline
batteries or
rechargeable NiCad
batteries
low perfusion:
Display type
Power supply
Battery capacity
blue monochrome
backlit LCD
4 x AA alkaline or
lithium batteries
19 hours typical
(alkaline)
40 hours typical
(lithium)
LED
4 x AA alkaline
batteries or
rechargeable NiMH
battery pack or
mains via external
charging stand
2500:
100 hours (alkaline)
45 hours (rechargeable)
30-254 bpm
8 second averaging
100 hours (maximum
display brightness)
approx. 24 hours in
continuous mode or
1500 spot checks
(1 min on/2 min off)
2500A:
60 hours (alkaline)
alarms (2500A only);
72 hour internal
memory; serial
interface & data
management software
200 hours (minimum
display brightness)
158 x 73 x 35 mm
138 x 70 x 32 mm
150 x 80 x 20 mm
spot check &
continuous modes;
1000 spot check,
99 patient internal
memory; RS232C
i li
f
172 x 84 x 37 mm
Weight
0.290 kg
approx. 0.210-0.230 kg
depending on model
0.280 kg
0.462 kg
£811.13
2500 £499.00
2500A £556.00
8500 £349.00
with memory £399.00
£265.00
range of sensors;
external printer;
carry case
range of sensors;
carry case
range of sensors;
thermal printer; carry
case; rubber bumper
disposable & reusable
sensor range;
carrying case; rubber
boot
Other features
Size (H x W x D)
Accessories &
prices
alarms; perfusion
value; infrared printer
interface
spot check &
continuous modes;
optional alarms &
memory
22
SpO2 technology
Sensor wavelengths
Smiths Medical
BCI FingerPrint
Smiths Medical
BCI MiniCorr
Smiths Medical
SPECTRO2 10
Smiths Medical
SPECTRO2 20
BCI
BCI SAC
(Serial Auotcorrelation)
BCI
BCI SAC (Serial
Auotcorrelation)
not specified
not specified
not specified
not specified
100%-0%;
8 beat averaging
SpO2 range &
accuracy
99%-0%
8 beat averaging
99%-70%: ±2%
<70%: unspecified
99%-0%
4, 8 or 16 beat
averaging
99%-70%: ±2%
69%-50%: ±3%
<50%: unspecified
99%-0%;
8 beat averaging
no motion or low
perfusion:
70%-100%:
±2 Arms (adult/paed);
±3 Arms (neo)
no motion or low
perfusion:
70%-99%: ±2 Arms
(adult/paed);
<70%: unspecified
motion:
70%-100%:
<70%: unspecified
Pulse rate range &
accuracy
30-254 bpm
8 second averaging
±2% or ±2 bpm
20-300 bpm;
8 second averaging
30-254 bpm
8 or 16 second
averaging
30-254 bpm;
8 second averaging
30-254 bpm ±2 Arms
±2 Arms
±2% or ±2 bpm
low perfusion:
30-250 bpm ±3 Arms
low perfusion:
25-250 bpm ±3 Arms
Display type
LED
LED
LED
Power supply
4 x AA alkaline or
rechargeable
equivalent batteries
6 x AA alkaline
batteries or
optional external
mains adaptor
4 x AA alkaline batteries, rechargeable lithium-ion
battery pack, external mains adaptor/charger or
external USB
Battery capacity
Other features
Size (H x W x D)
14 hours
(continuous printing) or
24 hours
32 hours (alkaline)
26 hours (alkaline)
24 hours (no printing)
(without printing)
54 hours (rechargeable) 31 hours (rechargeable)
or 2000 spot checks
integrated thermal
alarms; clinician, sleep
printer; 14 hour,
& home modes;
30 second interval,
12, 24 or 90 hour
pulse amplitude index;
99 patient internal
(4, 8 or 30 sec interval),
144 hour, 4 second
as for model 10 plus
trend memory; auto
99 patient internal
interval, 99 patient
Nellcor sensor
power off; protective
memory; protective
internal memory;
compatibility
rubber boot included;
rubber boot included ;
integrated sensor
PC interface;
infrared printer interface; storage; USB interface
FingerPrint Sleep
RS232 serial PC
version
interface
167 x 70 x 36 mm
167 x 70 x 36 mm
155 x 84 x 43 mm
Weight
0.369 kg
0.454 kg
0.340 kg
£285.00
£485.00
sensor range; carrying
case; printer paper
sensor range; carrying
case; mains adaptor
Accessories &
prices
price not provided
price not provided
disposable & reusable sensor range; docking
station; attachable printer; protective rubber boot
(4 colours); mounting bracket
23
SpO2 technology
Sensor wavelengths
Smiths Medical
SPECTRO2 30
SPO Medical
PulseOx 6100
Viamed
VM-2160
BCI SAC (Serial
Auotcorrelation)
SPO Medical
Reflective Pulse
Oximetry (RPO) with
AutoSpot™ technology
Viamed
not specified
not specified
red 660 nm
infrared 905 nm
100%-0%;
2 (sleep mode), 4, 8 or
16 beat averaging;
SpO2 range &
accuracy
no motion or low
perfusion:
70%-100%:
±3 Arms (neo)
99%-40%
99%-70%:
±2% or ±2 digits
<70%: unspecified
100%-0%
±1SD = 2%
Arms = <3%
motion:
70%-100%:
<70%: unspecified
Pulse rate range &
accuracy
20-300 bpm;
8 or 16 second
averaging
±2 Arms
40-250 bpm
20-300 bpm
±3% or ±3 digits
±1 digit (≤100 bpm)
±1% (>100 bpm)
low perfusion:
25-250 bpm ±3 Arms
Display type
LED
blue backlit
monochrome LCD
128 x 160 pixel
colour OLED
Power supply
4 x AA alkaline
batteries, rechargeable
lithium-ion battery pack,
or external USB
2 x AA alkaline
batteries
3 x AA alkaline
batteries
Battery capacity
17 hours (alkaline)
200 hours
as for model 20 plus
alarms & clinician,
sleep & home modes
and 72 hour,
2-30 second interval
internal memory
internal memory;
auto power off
155 x 84 x 43 mm
160 x 44 x 27 mm
118 x 60 x 25 mm
0.340 kg
0.085 kg
0.160 kg
price not provided
£245.00
£345.00
Accessories &
prices
as for SpectrO2 10
and SpectrO2 20
Other features
Size (H x W x D)
Weight
>2 days continuous or
approx. 5 days in
economy power mode
alarms; 15, 30 or 240
minute short term or up
to 480 hour, 50 dataset
internal memory;
protective cover; USB
2.0 interface & PC
software
carrying bag £14.00;
mounting options
24
Unpublished report
Non-peer reviewed
article
Peer-reviewed
article (comparison
with other pulse
oximeters)
Peer-reviewed
article (comparison
with co-oximetry)
Table 5. Evidence of accuracy and performance of handheld pulse oximeters
Manufacturer
Model
Beijing Choice
MD300K1-E
no evidence
Bionet
Oxy9
no evidence
Bitmos Medizintechnik
SAT 800
9
[13-15, 18]
9
[22, 24, 2731, 33-36]
SAT 801
9
[13-15, 18]
9
[22, 24, 2731, 33-36]
Comdek
MD-600P/MD-610P
no evidence
Comdek
MD-630P
no evidence
Comdek
MD-670P
no evidence
Comdek
MD-680P
no evidence
Criticare
503DX miniSPO2T
Daray
V60
no evidence
Daray
V202
no evidence
Edan
H100B
no evidence
GE Healthcare
TuffSat
9
[37]
Huntleigh
smartsigns MiniPulse
9
[37]
Masimo
Rad-5/Rad-5v
9
[13-15, 18]
9
[22, 24, 2731, 33-36]
Masimo
Rad-57
9
[13-15, 18]
9
[22, 24, 2731, 33-36]
Mediaid
M34
9
[44]
Mediaid
M5340
9
[44]
Mindray
PM-50
no evidence
Mindray
PM-60
no evidence
Nellcor
Oximax N-65
Comment
9
[9]
9
[48]
9
[50-55]
25
Unpublished report
Non-peer reviewed
article
Peer-reviewed
article (comparison
with other pulse
oximeters)
Peer-reviewed
article (comparison
with co-oximetry)
Table 5. Evidence of accuracy and performance of handheld pulse oximeters (continued)
Manufacturer
Model
Nonin
PalmSAT 2500/2500A
9
[17, 23]
9
[39-41]
Nonin
8500 series
9
[17, 23]
9
[39-41]
Smiths Medical
BCI 3301
9
[37]
Smiths Medical
BCI FingerPrint
9
[37]
Smiths Medical
BCI MiniCorr
9
[37]
Smiths Medical
SPECTRO2 10
9
[37]
Smiths Medical
SPECTRO2 20
9
[37]
Smiths Medical
SPECTRO2 30
9
[37]
SPO Medical
PulseOx 6100
Viamed
VM-2160
Comment
9
[41, 43]
no evidence
Wrist-worn pulse oximeters
26
Table 6. Wrist-worn pulse oximeters
Konica Minolta
Pulsox-300
Konica Minolta
Pulsox-300i
Nonin
3100 WristOx
SpO2 technology
Konica Minolta
Nonin PureSAT®
Sensor wavelengths
red 665 nm
infrared 880 nm
red 660 nm
infrared 910 nm
100%-0%;
5 second averaging
SpO2 range &
accuracy
100%-70%: ±1sd: ±2%
<70% unspecified
30-230 bpm
5 second averaging
Pulse rate range &
accuracy
30-100 bpm: ±2 bpm
100-230 bpm: ±2% of reading
Display type
LED
Power supply
Battery capacity
100%-0%
100%-70%: ±2 digits
<70% unspecified
backlit LCD
1 x AAA alkaline battery
approx. 16 hours
approx. 30 hours
alarms
Other features
Pulsox-300i only: up to 300 hour, 1 second interval internal
memory; USB interface and
PC analysis software with 0.1% resolution
Size (H x W x D)
Weight
Accessories &
prices
18-300 bpm
±3%
LCD
2 x size N alkaline batteries
at least 24 hours continuous
3 activation modes (spot check,
sensor activation &
programmed);
33, 16 or 8 hour
(4, 2 or 1 second interval)
internal memory;
3 minute auto power off;
serial interface with PC analysis
software
68 x 58 x 15 mm
51 x 44 x 19 mm
0.056 kg (including battery)
0.025 kg
(excluding battery)
£558.57
with finger probe
£350.00
without finger probe
£607.57
with finger probe
£399.00
without finger probe
£695.00
inc. reusable finger sensor
range of sensors
PC software with
USB adaptor & cable
set £399.00
range of sensors
PC software £379.00
serial interface
cable £89.70
carrying bag £19.95
range of sensors
Wrist-worn pulse oximeters
27
Unpublished report
Non-peer reviewed
article
Peer-reviewed
article (comparison
with other pulse
oximeters)
Peer-reviewed
article (comparison
with co-oximetry)
Table 7. Evidence of clinical study to confirm accuracy of wrist-worn pulse oximeters
Manufacturer
Model
Konica Minolta
Pulsox-300
9
[43, 46]
Konica Minolta
Pulsox-300i
9
[43, 46]
Nonin
3100 WristOx
9
[40, 43]
Comment
Bedside / desktop pulse oximeters
28
Table 8. Bedside / desktop pulse oximeters
Bitmos
Medizintechnik
SAT 805
Criticare
504DX
Daray
V450
GE Healthcare
3800
SpO2 technology
Masimo SET®
Criticare Systems Inc
(CSI) DOX™
Daray
GE TruTrak®+
Sensor wavelengths
red 660 nm
infrared 940 nm
not specified
red 660 nm
infrared 940 nm
red 650-670 nm
infrared 930-950 nm
100%-1%
SpO2 range &
accuracy
adults & children no motion:
±2 digits
neonates - no motion &
adults & children with motion:
±3 digits
99%-0%
99%-70%: ±2%
69%-40%: ±3%
<40% unspecified
100%-35%
100%-70%: ±2%
69%-0% unspecified
no motion:
100%-70%: ±2 digits
with motion:
100%-70%: ±3 digits
<70% unspecified
69%-0% unspecified
Pulse rate range &
accuracy
100%-0%
3, 6 or 12 second
averaging
25-240 bpm
20-300 bpm
no motion: ±3 digits
with motion: ±5 digits
±1 bpm or ±1%,
30-255 bpm
30-250 bpm
±2 bpm
±2 bpm or ±2%,
with motion: unspecified
Display type
backlit LCD
LED (numerics) &
LCD (text)
LED (numerics) &
blue monochrome
backlit LCD
(waveform & settings)
Power supply
internal rechargeable
7.2V lithium ion battery
with external mains
adaptor & charger
sealed lead acid
mains charger
lithium battery;
or 5 x AA alkaline or
8V sealed lead acid
battery with integrated
mains
15 hours (minimum)
18 hours (typical)
6 hours (typical)
15 hours (lithium)
>5 hours (AA)
approx. 5.5 hours
alarms & trends; 2 MB,
160 hour internal flash
memory card;
plethysmogram
waveform; analogue &
nurse call outputs;
infrared interface with
PC analysis software
alarms; 24 hour,
4 second interval
internal memory;
RS232-compatible
serial interface with
sleep study software
alarms & trends; 36 hour
internal memory; auto
power off;
plethysmogram
waveform; serial
interface with PC
analysis software;
temperature model
available (V450T)
alarms; 12 hour
internal memory;
perfusion index value;
plethysmogram
waveform; serial
interface
Size (H x W x D)
92 x 240 x 104 mm
144 x 178 x 122 mm
148 x 160 x 85 mm
94 x 244 x 225 mm
Weight
0.900 kg
(including battery)
1.520 kg
(inc. battery & printer)
0.420 kg
2.800 kg
£999.00
£995.00
£399.00
£449.00 (V450T)
£1534.00
integrated thermal
printer option;
sensor range
adult, paediatric &
neonate sensors
£85.00
sensor range
Battery capacity
Other features
Accessories &
prices
LED (numerics) &
orange monochrome
backlit LCD
29
Table 8. Bedside / desktop pulse oximeters (continued)
SpO2 technology
Sensor wavelengths
SpO2 range &
accuracy
GE Healthcare
3900/3900P
GE Healthcare
TruSat
GE TruTrak®+
Ohmeda TruSignal®
red 650-670 nm
infrared 930-950 nm
red 650-670 nm
infrared 930-950 nm
100%-0%
3, 6 or 12 second
averaging
100%-1%
10 second averaging
Masimo
Rad-8
Masimo SET®
upgradeable to Masimo
Masimo SET®
Rainbow SET®
red 660 nm
infrared 940 nm
100%-0%
2, 4, 8, 10, 12, 14 or 16 second averaging
no motion:
100%-70%: ±2 digits
no motion & low
perfusion:
100%-70%: ±2 digits
100%-70%: no motion & low perfusion: ±2%
with motion:
100%-70%: ±3 digits
with motion:
100%-70%: ±3 digits
80%-60%: no motion: ±3% (Radical-7); ±4%
(Rad-8) for LNOP® blue adhesive sensors
<70% unspecified
<70% unspecified
30-250 bpm
30-255 bpm
Pulse rate range &
accuracy
Masimo
Radical-7
±2 bpm or ±2%,
with motion: unspecified
no motion: ±2 bpm or
±2%, whichever is
greater
low perfusion: ±3 bpm
with motion: ±5 bpm
with motion & neonates: ±3%
25-240 bpm
no motion & low perfusion: ±3 bpm
Display type
LED (numerics) &
orange monochrome
backlit LCD
orange monochrome
backlit LCD
multi-colour or blue
monochrome TFT LCD
LED
Power supply
8V sealed lead acid
mains
12V NiMH battery with
external mains charger
NiMH battery with
integrated mains
sealed lead acid
mains
approx. 5.5 hours;
4 hours
(continuous printing)
30 hours;
20 hours (with trend
download (TD) option)
4 hours
(10 hour option)
up to 7 hours
Battery capacity
Other features
alarms & trends; 24 hour
alarms; perfusion index
handheld or docking
internal memory;
value; 20 minute auto
station configuration;
power off; TD option:
plethysmogram
48 hour, 4 second
alarms; display rotation;
waveform; analogue
interval trend memory &
72 hour, 2 second
output; serial interface
download with RS-232
interval trends; serial,
with PC software;
serial interface & PC
analogue, nurse call &
integrated thermal
software
interface options
printer model (3900P)
home & sleep modes;
alarms; 72 hour,
2 second interval
trends; serial, nurse
call & interface options
Size (H x W x D)
94 x 244 x 225 mm;
104 x 244 x 225 mm
(with printer)
103 x 218 x 115 mm
89 x 226 x 53 mm
267 x 89 x 196 mm
(in docking station)
76 x 208 x 152 mm
Weight
2.800 kg;
3.200 kg (with printer)
1.250 kg;
1.47 kg (with TD option)
0.540 kg (handheld)
up to 1.950 kg
0.908 kg
£1100.00 (blue)
£1600.00 (colour)
£800.00
reusable sensor range
Accessories &
prices
£1903.00
£1200.00
printer upgrade £826.00 £1350.00 with TD option
sensor range
sensor range;
printer kit £564.00
30
Masimo
Rad-87
SpO2 technology
Sensor wavelengths
SpO2 range &
accuracy
Masimo SET®
upgradeable to Masimo
Rainbow SET®
red 660 nm
infrared 940 nm
100%-0%
2, 4, 8, 10, 12, 14 or
16 second averaging
Nellcor
Oximax N-560
Noni PureSAT®
red 660 nm
infrared 900 nm
red 660 nm
infrared 910 nm
100%-1%
100%-70% (±1Arms =
68% of measurements):
no motion:
adult/paed: ±2 digits
neonate: ±3 digits
100%-0%
100%-70%: no motion
& low perfusion: ±2%
100%-70% (±1sd):
adult-neonate & low perfusion: ±2 digits
with motion &
neonates: ±3%
80%-60% (±1sd):
adult-neonate:
±3 digits
motion:
neonate: NA
20-250 bpm
no motion: 18-300 bpm
motion & low perfusion:
40-240 bpm
±3 digits
including low perfusion
no motion & low
25-240 bpm
no motion & low
perfusion: ±3 bpm
Nonin
7500/7500FO
Nellcor OxiMax® with LoSat™
80%-60%: no motion:
±3% for LNOP® blue
adhesive sensors
Pulse rate range &
accuracy
Nellcor
Oximax N-600x
with motion: ±5 digits
Display type
LED
LED
blue monochrome
backlit LCD
Power supply
sealed lead acid
mains
9.6V NiMH battery with
integrated mains
sealed lead acid
mains
7.2V NiMH battery with
external mains charger
up to 4 hours
8 hours minimum
7 hours typical
16 hours minimum
alarms; wireless network
communication;
72 hour, 2 second
interval trends; serial,
nurse call & interface
options
alarms; 24 hour trend;
nurse call; serial
output
fast averaging mode;
alarms; on-screen
24-48 hour trends;
nurse call; serial and
analogue outputs
7500FO (fibre optic
sensors) designed for
MRI use; alarms;
70 hour internal
memory; analogue
output; serial interface
with PC software
76 x 208 x 152 mm
75 x 230 x 128 mm
84 x 264 x 173 mm
92 x 219 x 142 mm
Weight
0.908 kg
1.390 kg
2.600 kg
approx. 0.900 kg
(inc. battery)
£1000.00
£1460.03
£2703.75
7500 £795.00
7500FO £1395.00
Battery capacity
Other features
Size (H x W x D)
Accessories &
prices
range of OxiMax sensors
LED
carry case £78.20
range of sensors
31
Nonin
Avant 4000
Nonin
Avant 9600
Nonin
Avant 9700
SpO2 technology
Noni PureSAT®
Sensor wavelengths
red 660 nm
infrared 910 nm
100%-0%
SpO2 range &
accuracy
100%-0%
100%-70%:
adults with finger clip
sensor: ±2 digits
adults with flex sensor:
±3 digits
100%-70% (±1sd):
motion, no motion & low perfusion:
neonates: ±3 digits
18-300 bpm
no motion 18-300 bpm
& low perfusion
Nonin
Medair
PulseSense
100%-0%
adult/paediatric:
±2 digits
18-300 bpm
Pulse rate range &
accuracy
±3%
motion 60-240 bpm:
±5 digits
Display type
LED
Power supply
main unit: internal
rechargeable 7.2V NiMH
battery pack with
wireless sensor:
2 x AA batteries
Battery capacity
Other features
main unit:
18 hours minimum
wireless sensor:
120 hours minimum
Bluetooth® wireless
wrist-worn sensor;
alarms; 33.5 hour
internal memory;
RS-232 serial interface
with PC software
Size (H x W x D)
Weight
18-255 bpm
±2 digits
low perfusion:
LED (numerics)
backlit colour LCD
(waveform)
touchscreen blue monochrome backlit LCD
internal rechargeable 7.2V NiMH battery pack
with external mains adaptor
lithium ion battery with
LED
12 hours minimum
12 hours
(LCD backlight off)
8 hours
(LCD backlight on)
alarms; 115 hour
as for 9600 with
internal memory; nurse additional colour-change
call; RS-232 serial
(green/amber/red)
interface with PC
plethysmogram
software
waveform
140 x 184 x 114 mm
approx. 12 hours
alarms; plethysmogram
waveform; 1.5 hour
internal trend memory
with on-screen display;
RS-232 serial interface
with PC software
135 x 200 x 50 mm
1.000 kg (main unit)
0.125 kg (sensor)
1.000 kg
1.100 kg
0.800 kg
£1595.00
£995.00
£1445.00
£850.00
Accessories &
prices
carry case £89.50
range of sensors
download kit £259.00
range of sensors
Smiths Medical
BCI 3180
Smiths Medical
BCI Autocorr
BCI
BCI SAC
(Serial Autocorrelation)
not specified
not specified
SpO2 technology
Sensor wavelengths
100%-0%
4, 8 or 16 beat averaging
SpO2 range &
accuracy
100%-70%:
adult/paed: ±2%
neonate: ±3%
<70% unspecified
100%-0%
4, 8 or 16 beat averaging
100%-70%:
no motion: ±2%
motion: ±3%
69%-50%:
no motion: ±3%
30-254 bpm
8 or 16 second
averaging
30-254 bpm
8 or 16 second
averaging
±2 bpm or ±2%,
no motion: ±2%
motion: unspecified
Display type
blue monochrome backlit
LCD
LED
Power supply
internal rechargeable 6V
sealed lead acid battery
with integrated mains
sealed lead acid battery
with external mains
adaptor
approx. 4 hours
approx. 4.5 hours
continuous
alarms; plethysmogram
waveform; 30 hour internal
trend memory with onscreen display; serial
infrared printer output
alarms; printer output;
digital, analogue & nurse
call outputs
165 x 203 x 127 mm
82 x 216 x 140 mm
Weight
2.000 kg
0.850 kg
£1042.36
£785.00
sensor range
sensor range
Pulse rate range &
accuracy
Battery capacity
Other features
Size (H x W x D)
Accessories &
prices
32
33
9
[22, 24,
27-31, 3336]
9
[9, 21]
9
[26]
Unpublished report
Peer-reviewed
article (comparison
with other pulse
oximeters)
9
[13-15, 18]
Non-peer reviewed
article
Peer-reviewed
article (comparison
with co-oximetry)
Table 9. Evidence of clinical study to confirm accuracy of bedside/desktop pulse oximeters
Manufacturer
Model
SAT 805
Criticare
504DX
Daray
V450
GE Healthcare
3800
9
[38]
GE Healthcare
3900/3900P
9
[38]
GE Healthcare
TruSat
Masimo
Radical-7
9
[13-16, 1820]
9
[22, 24,
27-36]
Masimo
Rad-8
9
[13-15, 18]
9
[22, 24,
27-31, 3336]
Masimo
Rad-87
9
[13-15, 18]
9
[22, 24,
27-31, 3336]
Nellcor
Oximax N-560
9
[50-55]
Nellcor
Oximax N-600x
9
[50-55]
Nonin
7500/7500FO
9
[40]
Nonin
Avant 4000
9
[40]
Nonin
Avant 9600
9
[40]
Nonin
Avant 9700
Nonin
Medair PulseSense
Smiths Medical
BCI 3180
9
[37]
Smiths Medical
BCI Autocorr
9
[37]
Comment
no evidence
9
[56, 57]
9
[47, 49-55]
9
[49]
9
[19]
9
[42]
9
[40, 42]
9
[40]
Choosing a pulse oximeter
34
The product information presented in this market review should be used with the
following questions to shortlist pulse oximeters that best meet purchasers
requirements. Under each question is a list of relevant features to consider.
1. In what setting is the pulse oximeter to be used (eg GP surgery, ambulance)?
• Display type
• Power supply
• Other features
• Size
• Weight
2. What are the lifetime costs of the device?
• Battery capacity
• List price
• Accessories and prices
3. How accurate is the pulse oximeter and does it comply with the BS EN ISO
9919:2009 [6]?
• SpO2 range and accuracy – table 10 lists the pulse oximeters in the market
review for which we found evidence of accuracy testing; for the level of
evidence see tables 3, 5, 7 and 9.
• Pulse rate range and accuracy
We recommend that only those devices with evidence of accuracy testing (table 10)
be considered for shortlisting.
Once a shortlist of devices has been created, we recommend that demonstrations or
short term loans are arranged with suppliers. The loan period should be used to
assess qualities that have not been considered in this market review, such as
usability, build quality and ergonomic features.
35
Bitmos Medizintechnik SAT 801
9
9
Criticare 503DX miniSPO2T
9
Criticare 504DX
9
GE Healthcare 3800
9
GE Healthcare 3900/3900P
9
GE Healthcare TuffSat
9
9
GE Healthcare TruSat
Huntleigh smartsigns MiniPulse
Wrist-worn
9
Bedside/desktop
Fingertip
Device
Handheld
Table 10. Pulse oximeters with evidence of accuracy and performance testing
9
Konica Minolta Pulsox-1
9
9
9
Konica Minolta Pulsox-300i
9
Masimo Rad-5/Rad-5v
9
Masimo Rad-57
9
Masimo Radical-7
9
Masimo Rad-8
9
Masimo Rad-87
9
9
Mediaid 100/100C
Mediaid M34
9
Mediaid M5340
9
Nellcor Oximax N-65
9
Nellcor Oximax N-560
9
Nellcor Oximax N-600x
9
Nonin PalmSAT 2500/2500A
9
Nonin 3100 WristOx
9
Nonin 7500/7500FO
Nonin 8500 series
9
9
Nonin 9500 Onyx
9
Nonin 9550 Onyx II
9
Nonin 9560 Onyx II
9
Nonin Avant 4000
9
36
Nonin Avant 9600
9
Nonin Avant 9700
9
Nonin GO2
9
Nonin Medair OxyCheck
9
Nonin Medair PulseSense
9
Smiths Medical BCI 3180
9
Smiths Medical BCI 3301
9
Smiths Medical BCI Autocorr
9
9
Smiths Medical BCI Digit
Smiths Medical BCI FingerPrint
9
Smiths Medical BCI MiniCorr
9
Smiths Medical SPECTRO2 10
9
9
9
SPO Medical PulseOx 5500
9
9
9
Wrist-worn
Bedside/desktop
Fingertip
Device
Handheld
Table 10. Pulse oximeters with evidence of accuracy testing (continued)
Acknowledgements
We should like to thank the following for their contribution to this market review.
Manufacturers and suppliers
John Allen, Lead Clinical Scientist, Microvascular Diagnostics, Freeman Hospital,
Newcastle upon Tyne
37
Glossary
38
ARMS
accuracy of pulse oximeter stated in terms of the RMS
difference between measured SpO2 values and SaO2
reference values
bpm
beats per minute
CO2
carbon dioxide
ECG
electrocardiograph
FIO2
fraction of inspired oxygen
HbO2%
percentage of oxyhaemoglobin
HR
heart rate
Hypobaric
below normal pressure (applied to gases less than
atmospheric pressure)
ICU
intensive care unit
LCD
liquid crystal display
LED
light emitting diode
NHS
National Health Service
NICU
neonatal intensive care unit
nm
nanometre
Normoxia
normal levels of oxygen
OLED
organic light emitting diode
OT
operating theatre
PACU
postanesthesia care unit
PC
personal computer
PCO2
partial pressure of carbon dioxide
pH
a measure of the acidity of a solution
PO2 or PaO2
partial pressure of oxygen
PR
pulse rate
PTcO2
transcutaneous partial pressure of oxygen
R/IR
red to infrared ratio
RMS
root mean squared
SaCO
arterial carbon monoxide saturation
SaO2
arterial oxygen saturation
SD
standard deviation
Glossary
39
SpCO
carbon monoxide saturation measured non-invasively,
which is an approximation of SaCO
SpO2
oxygen saturation (via pulse oximeter); SpO2 is an
approximation of SaO2 when oxygen saturation is good
(generally less that 3% discrepancy when SpO2 is above
70% [58])
SET
signal extraction technology
USB
universal serial bus
VO2max
maximum amount of oxygen in millilitres used in one minute
per kilogram of body weight
References
40
1.
Reay CA, Menes JA, Bousfield DR, Colechin ES and Sims AJ. CEP
08018 Buyers' guide: Hospital grade non-invasive blood pressure monitors.
www.dh.gov.uk/cep, 2008.
2.
Nayyar P, Batki AD and Thomason H. CEP09043 Buyers' guide: Blood gas
analysers. www.dh.gov.uk/cep, 2009.
3.
Thomason HL, Batki AD and Nayyar P. CEP09044 Market review: Blood gas
analysers. www.dh.gov.uk/cep, 2009.
4.
Colechin ES, Bousfield DR, Reay CJ and Sims AJ. CEP10065 Buyers' guide:
Pulse oximeters in primary and prehospital care. www.dh.gov.uk/cep, 2010.
5.
Colechin ES, Reay CA, Bousfield DR and Sims AJ. CEP10064 Evidence
review: Pulse oximeters in primary and prehospital care. www.dh.gov.uk/cep,
2010.
6.
British Standards Institute. BS EN ISO 9919: Medical electrical equipment Particular requirements for the basic safety and essential performance of
pulse oximeter equipment for medical use. 2009.
7.
International Organisation for Standardization. ISO 9919: Pulse oximeters for
medical use - Requirements. 1992.
8.
British Standards Institute. BS EN ISO 9919: Medical electrical equipment Particular requirements for the basic safety and essential performance of
pulse oximeter equipment for medical use. 2005.
9.
Clayton DG, Webb RK, Ralston AC, Duthie D and Runciman WB. A
comparison of the performance of 20 pulse oximeters under conditions of poor
perfusion. Anaesth 1991 46(1):3-10.
10.
Trivedi NS, Ghouri AF, Shah NK, Lai E and Barker SJ. Effects of motion,
ambient light, and hypoperfusion on pulse oximeter function. J Clin Anesth
1997 9(3):179-83.
11.
Wood RJ, Gore CJ, Hahn AG, Norton KI, Scroop GC, Campbell DP, Watson
DB and Emonson DL. Accuracy of two pulse oximeters during maximal cycling
exercise. Aust J Sci Med Sport 1997 29(2):47-50.
12.
Chiappini F, Fuso L and Pistelli R. Accuracy of a pulse oximeter in the
measurement of oxyhaemoglobin saturation. Eur Respir J 1998 11(3):716-9.
References
41
13.
Bohnhorst B, Peter CS and Poets CF. Detection of hyperoxaemia in neonates:
data from three new pulse oximeters. Arch Dis Child Fetal Neonatal Ed 2002
87(3):F217-9.
14.
Durbin CGJ and Rostow SK. More reliable oximetry reduces the frequency of
arterial blood gas analyses and hastens oxygen weaning after cardiac
surgery: a prospective, randomized trial of the clinical impact of a new
technology. Crit Care Med 2002 30(8):1735-40.
15.
Durbin CGJ and Rostow SK. Advantages of new technology pulse oximetry
with adults in extremis. Anesth Analg 2002 94(1 Suppl):S81-3.
16.
Torres AJ, Skender KM, Wohrley JD, Aldag JC, Raff GW, Bysani GK and
Geiss DM. Pulse oximetry in children with congenital heart disease: effects of
cardiopulmonary bypass and cyanosis. J Intensive Care Med 2004
19(4):229-34.
17.
Bickler PE, Feiner JR and Severinghaus JW. Effects of skin pigmentation on
pulse oximeter accuracy at low saturations. Anesthesiol 2005 102(4):715-9.
18.
Murthy TVSP, Goyal R and Singh VP. Masimo - a new reliable non invasive
method of detecting oxygen saturation in critically ill. Indian J Anaesth 2005
49(2):133-6.
19.
Feiner JR, Severinghaus JW and Bickler PE. Dark skin decreases the
accuracy of pulse oximeters at low oxygen saturation: the effects of oximeter
probe type and gender. Anesth Analg 2007 105(6 Suppl):S18-23.
20.
Levrat Q, Petitpas F, Bouche G, Debaene B and Mimoz O. Usefulness of
pulse oximetry using the SET technology in critically ill adult patients. Ann Fr
Anesth Reanim 2009 28(7-8):640-4.
21.
Iyer P, McDougall P, Loughnan P, Mee RB, Al-Tawil K and Carlin J. Accuracy
of pulse oximetry in hypothermic neonates and infants undergoing cardiac
surgery. Crit Care Med 1996 24(3):507-11.
22.
Dumas C, Wahr JA and Tremper KK. Clinical evaluation of a prototype motion
artifact resistant pulse oximeter in the recovery room. Anesth Analg 1996
83(2):269-72.
23.
Macnab AJ, Smith M, Phillips N and Smart P. Oximeter reliability in a subzero
environment. Aviat Space Environ Med 1996 67(11):1053-6.
24.
Barker SJ and Shah NK. The effects of motion on the performance of pulse
oximeters in volunteers (revised publication). Anesthesiol 1997 86(1):101-8.
References
42
25.
Grieve SH, McIntosh N and Laing IA. Comparison of two different pulse
oximeters in monitoring preterm infants. Crit Care Med 1997 25(12):2051-4.
26.
Rheineck-Leyssius AT and Kalkman CJ. Advanced pulse oximeter signal
processing technology compared to simple averaging. I. Effect on frequency of
alarms in the operating room. J Clin Anesth 1999 11(3):192-5.
27.
Bohnhorst B, Peter CS and Poets CF. Pulse oximeters' reliability in detecting
hypoxemia and bradycardia: comparison between a conventional and two new
generation oximeters. Crit Care Med 2000 28(5):1565-8.
28.
Malviya S, Reynolds PI, Voepel-Lewis T, Siewert M, Watson D, Tait AR and
Tremper K. False alarms and sensitivity of conventional pulse oximetry versus
the Masimo SET technology in the pediatric postanethesia care unit. Anesth
Analg 2000 90(6):1336-40.
29.
Barker SJ. "Motion-resistant" pulse oximetry: a comparison of new and old
models. Anesth Analg 2002 95(4):967-72.
30.
Gehring H, Hornberger C, Matz H, Konecny E and Schmucker P. The effects
of motion artifact and low perfusion on the performance of a new generation of
pulse oximeters in volunteers undergoing hypoxemia. Respir Care 2002
47(1):48-60.
31.
Hay WWJ, Rodden DJ, Collins SM, Melara DL, Hale KA and Fashaw LM.
Reliability of conventional and new pulse oximetry in neonatal patients. J
Perinatol 2002 22(5):360-6.
32.
Irita K, Kain Y, Akiyoshi K, Tanaka Y and Takahashi S. Performance
evaluation of a new pulse oximeter during mild hypothermic cardiopulmonary
bypass. Anesth Analg 2003 96(1):11-4.
33.
Sahni R, Gupta A, Ohira-Kist K and Rosen TS. Motion resistant pulse oximetry
in neonates. Arch Dis Child Fetal Neonatal Ed 2003 88(6):F505-8.
34.
Robertson FA and Hoffman GM. Clinical evaluation of the effects of signal
integrity and saturation on data availability and accuracy of Masimo SET and
Nellcor N-395 oximeters in children. Anesth Analg 2004 98(3):617-22.
35.
Workie FA, Rais-Bahrami K and Short B. Clinical use of new-generation pulse
oximeters in the neonatal intensive care unit. Am J Perinatol 2005
22(7):357-60.
36.
Richards NM, Giuliano KK and Jones PG. A prospective comparison of 3 newgeneration pulse oximetry devices during ambulation after open heart surgery.
Respir Care 2006 51(1):29-35.
References
43
37.
Mathes AM, Kreuer S, Schneider SO, Ziegeler S and Grundmann U. The
performance of six oximeters in the environment of neuronavigation. Anesth
Analg 2008 107(2):541-4.
38.
Anonymous. Datex-Ohmeda TruTrak+ Oximetry Technology; A 663 patient,
multi-site performance evaluation during clinical motion. GE Healthcare.
http://www.gehealthcare.com/usen/oximetry/docs/OS3959.pdf 2002.
39.
Severinghaus JW. SpO2 accuracy of PureSAT signal processing technology The ONYX II. University of California at San Francisco Hypoxia Research
Laboratory. http://www.nonin.com/documents/TechnicalBulletin-ONYXII.pdf
2005.
40.
Nonin. Accuracy and superior performance of Nonin PureSAT and PureLight
oximetry technologies. University of California San Francisco Hypoxia
Research Laboratory.
http://www.nonin.com/documents/Technical_Bulletin_OEM_III.pdf , 2006.
41.
Nonin. Accuracy of Nonin Onyx 9500 fingertip pulse oximeter is superior to
competitor's device. University of California San Francisco Hypoxia Research
Laboratory. http://www.nonin.com/documents/Avant%2096009700%20Study%20Data.pdf 2007.
42.
University of California San Francisco. The effect of dark skin pigmentation
and low saturation in oximetry. University of California San Francisco.
http://www.nonin.com/documents/M5407%20Skin%20Pigmentation%20II%20Tech%20Bulletin.pdf , 2008.
43.
University of California San Francisco Hypoxia Research Laboratory. Nonin
PureSAT oximetry technlogy shown to have superior accuracy. University of
California San Francisco. http://www.nonin.com/documents/M6208%20WristOx%20Comparison%20Tech%20Bulletin.pdf , 2009.
44.
Mediaid. Mediaid Oximetry Technology. 2006.
45.
Feiner JR, Bickler PE and Severinghaus JW. Pulse oximeter report for Minolta
- Pulseox-2. Hypoxia Research Laboratory, University of California, San
Francisco. 2002.
46.
Feiner JR, Bickler PE and Severinghaus JW. Pulse oximeter report for Konica
Minolta Sensing, Inc. - P300i. Hypoxia Research Laboratory, University of
California, San Francisco. 2005.
47.
Goldstein MR, Louie N, Lawas-Alejo P, Pernia ML and Martin GI. Comparison
of two new generation pulse oximeters: The Masimo Radical and the Philips
Viridia. Anesthesiol 2003 99:A554.
References
44
48.
Mottram C, Hanson LJ and Scanlon PD. Comparison of Masimo Rad-57 pulse
oximeter with SpCO technology against a laboratory CO-oximeter using
arterial blood. Respir Care 2005 50(11):1471.
49.
Shah N, Taleghani A, Chitkara A and Miller JM. A comparison of accuracy of
three new generation pulse oximeters (POs) during motion and low perfusion
in human volunteers. Anesthesiol 2005 103:A1168.
50.
Ahrens TS and Ott K. Comparison of three new generation pulse oximeters in
a medical intensive care unit. Barnes Jewish Hospital.
http://www.nellcor.com/clin/List.aspx?S1=CLN&S2=POX 2006.
51.
Sautel M. Comparison of two newer-generation pulse oximeters in the
neonatal intensive care unit (NICU). Children's Memorial Hospital.
52.
Shah N and Estanol L. Impact of motion and low perfusion on SpO2 and pulse
rate in three new generations of POs in volunteers. Anesthesiol 2006
105:A1433.
53.
Shah N and Estanol L. Comparison of three new generation pulse oximeters
during motion and low perfusion in volunteers. Anesthesiol 2006 105:A929.
54.
Cox PN. New pulse oximetry sensor with low saturation accuracy claims - a
clinical evaluation. Anesthesiol 2007 107:A1540.
55.
Lussky RC, Cifuentes RF, Kaester K, Stanek G and Geppert J. A comparison
of newer generation pulse oximeters for accuracy and drop-out rate in
neonatal intensive care. Hennepin County Medical Center.
56.
Raley DM. Results of SpO2 accuracy of Datex-Ohmeda TruSatTM pulse
oximeter with OxyTip+ sensors during low perfusion conditions as compared
to co-oximetry. Datex-Ohmeda Inc. 2004.
57.
Raley DM. Results of oxygen saturation test of TruSatTM pulse oximeter during
motion and non-motion conditions compared to arterial blood co-oximetry.
Datex-Ohmeda. 2004.
58.
Design of pulse oximeters. Webster JG (Ed.). Bristol: Institute of Physics
Publishing Ltd, 1997.
59.
Barker SJ and Shah NK. Effects of motion on the performance of pulse
oximeters in volunteers. Anesthesiol 1996 85(4):774-81.
Appendix 1: Supplier contact details
Charter Kontron
Avant Business Centre
21 Denbigh Road
Denbigh West
Milton Keynes
MK1 1DT
45
(Supplies: Bitmos Medizintechnik)
Tel: 01908 646070
Fax: 01908 646030
website: www.charter-kontron.com
email: [email protected]
Covidien
Respiratory and Monitoring Solutions
154 Fareham Road
Gosport
Hampshire
PO12 0AS
(Supplies: Nellcor)
Tel: 01329 224114
Fax: 01392 224336
website: www.nellcor.com
Daray Ltd
Marquis Drive
Moira
Swadlincote
Derbyshire
DE12 6EJ
Tel: 0800 804 8384
Fax: 0333 321 0974
website: www.daray.co.uk
DeVilbiss Healthcare
Sunrise Medical Ltd
High Street
Wollaston
Stourbridge
West Midlands
DY8 4PS
Dolby Medical Ltd
Monitor House
Kerse Road
Stirling
FK7 7RZ
(Supplies: Konica Minolta)
Tel: 01384 446868
Fax: 01384 446628
website: www.devilbisshealthcare.com
(Supplies: Criticare Systems Inc)
Tel: 0800 111 4415
Fax: 01786 446630
website: www.dolbymedical.com
46
GE Healthcare
Clinical Systems Devices
71 Great North Road
Hatfield
Hertfordshire
AL9 5EN
Tel: 01707 263570
Fax: 01707 260065
website: www.gehealthcare.com/uken/
email: via form on website
Huntleigh Healthcare Ltd
Diagnostic Products Division
35 Portmanmoor Road
Cardiff
CF24 5HN
Inspiration Homecare Ltd
Gildor House
West Street
East Shilton
Leicester
LE9 7EJ
Tel: 029 2048 5885
Fax: 029 2049 2520
website: www.huntleigh-healthcare.com
(Supplies: Comdek and SeQual)
Tel: 01455 849187
Fax: 01455 846789
website: www.inspiration-homecare.co.uk
Masimo
Unit Q
Lodden Business Centre
Roentgen Road
Basingstoke
Hampshire
RG24 8NG
Tel: 01256 479988
Fax: 01256 473684
website: www.masimo.co.uk
Medical Physics International Ltd
Bolton Technology Exchange
Spa Road
Bolton
BL1 4AY
(Supplies: Bionet)
Tel: 01204 555900
Fax: 01204 555909
website: www.medphysics.co.uk
47
Mindray (UK) Ltd
3 Percy Street
St John’s Park
Huntingdon
Cambridgeshire
PE29 6SZ
PMS (Instruments) Ltd
Waldeck House
Waldeck Road
Maidenhead
Berkshire
SL6 8BR
Tel: 01480 416840
Fax: 01480 436588
website: www.mindray.com
email: unknown
(Supplies: Mediaid and RossMax)
Tel: 01628 773233
Fax: 01628 770562
website: www.pmsinstruments.co.uk
Proact Medical Ltd
9-13 Oakley Hay Lodge
Great Folds Road
Great Oakley
Northamptonshire
NN18 9AS
(Supplies: Nonin)
Tel: 0845 051 4244
Fax: 0845 051 4255
website: www.proactmedical.co.uk
Pulmolink
Redwood House
Canterbury Road
Charing
Kent
TN27 0EU
Pulse Medical Ltd
West Lea
Fosters Lane
Knaphill
Woking
Surrey
GU21 2PW
(Supplies: SPO Medical)
Tel: 01233 713070
Fax: 01233 713859
website: www.pulmolink.co.uk
(Supplies: Edan Instruments)
Tel: 01483 473803
Fax: 01483 326025
website: www.pulsemedical.co.uk
48
Smiths Medical International (UK) Ltd
Colonial Way
Watford
Hertfordshire
WD2 4LG
Viamed Ltd
15 Station Road
Cross Hills
Keighley
West Yorkshire
BD20 7DT
Tel: 01923 246434
Fax: 01923 231595
website: www.smiths-medical.com/uk/
(Supplies Viamed and Beijing Choice
MD300-C, MD300-C5 and MD300-D)
Tel: 01535 634542
Fax: 01535 635582
website: www.viamed.co.uk
White Medical
Sir Frank Whittle Business Centre
Great Central Way
Butlers Leap
Rugby
CV21 3XH
Williams Medical Supplies Ltd
The Maerdy Industrial Estate
Rhymney
Gwent
NP22 5PY
(Supplies: Nissei)
Tel: 01788 553904
Fax: 01788 560820
website: www.white-medical.co.uk
(Supplies: Mindray and Beijing Choice
M-Pulse,M-Pulse Lite and MD300K1-E)
Tel: 01685 844739
Fax: 01685 844725
website: www.wms.co.uk
Appendix 2: Summary of clinical trials
49
Table 11. Summary of peer-reviewed clinical trials
Authors
Study design
Outcomes
Results
Limitations
Source of funding
Clayton et al
1991 [9]
• Population: 120 patients, who were sedated and
undergoing intermittent positive pressure ventilation.
• Accuracy and
precision of each
pulse oximeter in
respect to cooximeter.
• Most oximeters failed to give a reading on a
number of patients as a result of poor signal
quality.
• Oximeters were
not tested on all
120 patients.
Not specified.
• Aim: To compare the performance of 20 oximeters
with in vitro measurements of haemoglobin
saturation with a co-oximeter in patients after cardiac
surgery involving cardiopulmonary bypass and
hypothermia.
• Method:
-
Patients were studied for 30 minutes to 2 hours.
-
Age, sex, operation, rectal temperature, systolic,
diastolic and mean blood pressure and drug
therapies of each patient were recorded.
-
Arterial pressures were monitored directly and
continuously through radial artery cannulae.
-
Haemoglobin saturation and PR displayed on
oximeters were recorded. Sample of arterial
blood was taken from radial artery cannula. It
was noted when pulse oximeters failed to give
readings of saturation.
-
-
Oximeters tested included: Datex Satlite, Invivo
4500, Nellcor N-200, Novametrix 505, Ohmeda
Biox 3740, Radiometer Oximeter, Datascope
Accusat, Ohmeda Biox 3700, Nonin 8604D,
Physio-Control 1600, Sensormedics Oxyshuttle,
Simed S-100, Spectramed Pulsat, Biochem
Microspan 3040, Criticare CSI 503, Criticare
CSI 504, Engstrom Eos, Kontron 7840, Minoltan
Pulsox 7 and Pulsemate Colin BX-5.
Oximeters were split into 3 groups and each
patient was test with all the oximeters in one
group. Each oximeter was tested on 40 different
patients.
• 5 of the 20 oximeters tested gave readings on all
40 patients.
• The number of readings within 2% of the cooximeter ranged from 11 to 32, while the number
of readings with 3% ranged from 20 to 40.
• The mean difference of the pulse oximeter
differed by 0.1 to 4.5% of co-oximeter saturation
measurement.
• 16 oximeters tended to overestimate saturation
whereas 4 underestimated.
• The precision of the oximeters varied from 0.96%
to 5.78%.
• Only 2 of the 20 oximeters met the criterion of
being within 1.96 SD of the true value.
• Under conditions of poor perfusion only 2
oximeters (Datex Satlite and Criticare CSI 503)
gave readings within 4% of reference cooximeter.
• 10 of the 20 oximeters failed to give readings at
least 10% of the time.
• Biochem Microspan ranked number 1 in terms of
accuracy but number 18 in terms of precision.
• Kontron 7840 ranked number 1 in terms of
precision but number 19 in terms of accuracy.
50
Table 11. Summary of peer-reviewed clinical trials (continued)
Authors
Study design
Outcomes
Results
Limitations
Source of funding
Dumas et al
1996 [22]
• Population: 50 patients who were generating
alarms with their routine pulse oximeter monitoring.
• Artefact effect on
measurements.
• 10 patients exhibited nonrhythmic motion.
Masimo SET device failed to give a signal 3
times for an average of 10 ± 0.6 seconds,
whereas the N-200 failed 14 times with a mean
duration of 46 ± 66 seconds.
• Small population.
Not specified.
• On 12 occasions the N-200 measured SpO2 to
be 10% lower than the Masimo SET SpO2.
During these episodes, the N-200 PR abruptly
changed and/or did not agree with the ECG HR.
• Not all SpO2
measurements
were compared
against blood gas
analysis.
• Aim: To compare a prototype Masimo SET device
with a conventional pulse oximeter in a
postanesthesia care unit (PACU).
• Method:
-
Conventional pulse oximeter was the Nellcor N200.
-
SpO2 and PR were recorded from both device
every second using a computer.
-
7 patients had arterial catheters to allow arterial
blood sampling for blood gas analysis.
-
Accuracy was measured by SpO2 values
obtained during brief disturbances compared
with baseline values and comparison of PR with
HR measured on ECG or via arterial catheter.
• During episodes of tremor exhibited in 5 patients,
26 occasions were observed where N-200 SpO2
was at least 10% lower than Masimo SET SpO2
and these disturbances persisted for 75 ± 105
seconds.
• During artefactual desaturations related to
patient motion the Masimo SET PR remained
consistent with the reference HR, whereas the
N-200 PR had a much larger bias for the 3 types
of motion.
• Low perfusion due to fist clenching caused N200 signal to be lost on 25 occasions compared
with 9 for the Masimo SET.
• Cool extremities resulted in complete signal loss
in N-200. The Masimo SET device did give
readings by with significant bias relative to
arterial blood gas measurements.
• The N-200 alarm frequency was once every 13
minutes, whereas the alarm frequency of the
Masimo SET was a factor of 2 lower.
• Comparison
mainly between
the 2 pulse
oximeters.
51
Authors
Study design
Outcomes
Results
Limitations
Source of funding
Iyer et al
1996 [21]
• Population: 25 infants undergoing major heart
surgery under hypothermic cardiopulmonary bypass.
• Effect on
accuracy of
temperature,
probe location,
haemoglobin
type and
medication.
• 151 observations were made using a hand probe
and 148 observations were made using a foot
probe.
Not specified.
• The pulse oximeter did not give a reading in 14
of the 151 instances and in 8 of the 148
instances.
• Study focused on
affects on pulse
oximeter accuracy
more than the
actual accuracy of
pulse oximeter.
• Aim: To assess the accuracy of pulse oximetry in
newborn babies and young infants recovering from
mild to profound hypothermia after open-heart
surgery and to determine the relationships of skin
temperature and probe site to the accuracy of the
pulse oximeter.
• Method:
-
Pulse oximetry readings were compared with
simultaneous, directly measured arterial blood
oxygen saturation using a hemoximeter. The
pulse oximeter used was the Criticare 504.
-
All blood samples were collected anaerobically,
stored on ice, and analysed within 0.5 hours of
collection. Pulse oximeter readings at the time
of sampling were collected.
-
Other observations recorded were peripheral
temperature at probe site, core temperature,
mean arterial blood pressure and details of
medication administered.
• 3 of the oximeter readouts with the foot probe
and 2 readings with the hand probe were <70%,
so were exclude from analysis.
• Temperature was shown to affect the accuracy
of pulse oximetry measurements for both the foot
and hand. When the foot temperature exceeded
29°C, 97.4% of observations were within the
acceptable range of ±3%, whereas the difference
between the pulse oximeter and hemoximeter
tended to be much greater below this
temperature. Similar observations were made
with the hand. At very low peripheral
temperatures, the proportion of oximeter
readings with inacceptable bias was greatly
increased at hand temperatures between 22°C
and 27°C.
• There was no significant difference between the
two sites in terms of dropout rate or in terms of
the proportion of readings with an abnormal bias.
• Increase in carboxyhemoglobin or
methemoglobin was not found to affect pulse
oximetry measurements.
• The administration of phenoxybenzamine was
not found to statistically affect SpO2 value preand post-injection.
• Paediatric
patients.
52
Authors
Study design
Outcomes
Results
Limitations
Source of funding
Macnab et al
1996 [23]
• Population: 26 children who were physiologically
stable and normathermic in a intensive care unit.
• Effect of device
temperature on
measurement.
• 26 pairs of SpO2 readings from Propaq
(temperature ≤-15°C at mid-point and ≤-30°C at
bottom) and control unit were obtained. There
was a significant correlation between the
readings and an analysis of variance revealed no
difference in mean levels. There was no
difference evident in ECG reading from Propaq
and control unit.
Not specified.
• Aim: To determine whether any of 3 commerciallyavailable, battery-operated, portable pulse
oximeters, used to monitor oxygen saturation during
transport, provide reliable data for oxygen saturation
and pulse rate in subzero environmental
temperatures comparable to those encountered by
inter-facility transport teams.
• Method:
-
The 3 oximeters tested were Propaq 106EL
monitor, Siemens Micro O2 oximeter and Nonin
8500N oximeter. They were tested against a
control unit, a Hewlett Packard Monitor
M1166A.
-
To simulate subzero environment, the
oximeters were packed in dry ice in a heavy
duty polystyrene cooler.
-
The approximate internal temperature of each
oximeter was calculated by measuring the
temperature at the mid zone of the cooler at the
surface of the oximeter screen and at the
bottom of the cooler.
-
The oximeters were left in the dry ice for at
least an hour prior to taking any SpO2
measurements.
-
Readings from the oximeters were compared
with simultaneous readings from the control
unit, which was at room temperature. Data was
collected immediately following simultaneous
measurements using control unit and arterial
blood gas sample.
-
Control unit data were considered acceptable
only if SpO2 value differed by 2% or less from
co-oximetry measurement.
• The Siemens oximeter repeatedly failed to give
readings under the hypothermic test conditions
and therefore no comparisons of data were
made.
• 18 pairs of SpO2 readings from Nonin
(temperature ≤-15°C at mid-point and ≤-30°C at
bottom) and control unit were obtained. There
was a significant correlation between the reading
and an analysis of variance revealed no
difference in mean levels. PR measurement
compared with the PR measurements of the
control unit.
• Paediatric
patients.
• Comparison
between pulse
oximeters and not
with blood gases
or co-oximetry
(gold standard).
• Did not test
devices in normal
operating
conditions.
53
Authors
Study design
Outcomes
Results
Limitations
Source of funding
Barker and
Shah 1997
[24]
• Population: 10 healthy volunteers.
• Accuracy,
dropout rate and
performance
index (PI).
• N-200 underestimated saturation by 5% to 18%
during motion.
Not specified.
(Revised
publication
of Barker
and Shah
1996 [59])
• Aim: To compare the performance during
standardised motion of three pulse oximeters:
Nellcor N-200, Nellcor N-3000 and Masimo SET.
• Method:
-
Each participant had a cannula placed in the
radial artery of their nondominant wrist. The
three oximeter sensors were attached to the test
hand (opposite hand to radial cannula).
• N-3000 failed to display SpO2 after a disconnectreconnection for both motions.
• Masimo SET displayed SpO2 at all times but
underestimated saturation by 3% to 6% during
hypoxemia and motion,
-
In addition N-200 and Masimo sensors were
placed on the cannulated hand.
• Masimo SET tracked the control SpO2 well
during desaturation but exhibited some lag
during rapid resaturation. The other two
oximeters did not tract control SpO2 in this case.
-
All sensors were provided by manufacturers of
oximeters.
• The controlled hand motion had significant effect
on accuracy and dropout rate of pulse oximeters.
-
Data obtained from the test hand during various
motion protocols were compared with
simultaneous data from a similar oximeter on
the stationary cannulated hand.
-
Standardised motions called rubbing and
tapping were generated by a motor-driven tilt
table. Various motion amplitude and frequencies
were tested.
• The test condition of connecting the oximeters
after beginning the motion proved more
strenuous than that of starting motion after
oximeters were connected and functioning. This
incurred a higher dropout rate and lower
performance index in the N-200 and N-3000.
-
Arterial blood was drawn from the radial cannula
at each stead-state condition of the study. Each
sample was processed using a blood gas
analyser for pH, PCO2 and PO2 and a cooximeter for HbO2%.
-
For hypoxaemia portions, participants breathed
through a tight-fitting face mask connected to an
anaesthesia machine, which was used to adjust
the inspired oxygen fraction. Oxygen saturations
were varied between 100% and 75%.
• Similar results were seen with rapid desaturation
to 75%, the N-3000 exhibited a higher error rate
and lower dropout rate.
• Masimo SET exhibited much lower error rates
and dropout rates that the other two oximeters
during motions.
• The lowest PI for Masimo was 95% compared
with 46% for N-3000 and 68% for N-200.
• For the alarm condition of SpO2 <90%, there
were no false negatives so the sensitivity for all
devices was 100%. Specificity of 70% and 80%
were reported for N-200 and N-3000.
• Though data was
measured,
comparison to
blood gases and
co-oximetry was
not made.
54
Authors
Study design
Outcomes
Results
Limitations
Source of funding
Grieve et al
1997 [25]
• Population: 16 infants in a neonatal unit.
• Comparison of
Nellcor and
Ohmeda
measurements.
• Nellcor pulse oximetry was artefactual 19% of
the total recorded time compared to 23% with
Ohmeda oximetry. All these episodes were
ascribed to infant movement.
Not specified.
• Aim: To test the reliability and variation in the
readings of Nellcor N-20 and Ohmeda Biox 3700.
• Method:
-
Pulse oximeters were used simultaneously, with
probes located on each foot.
-
ECG and transcutaneous oxygen tensions were
monitored continuously by a multiparameter
monitor.
-
All data was collected on one database.
-
Infants were observed while they were asleep,
awake and whilst feeding.
-
Artefact on pulse oximetry tracings were
determined by clinical observation, comparison
with simultaneous transcutaneous PO2 values
and by correlation of PR from pulse oximeters
with ECG HR.
• Nellcor pulse oximeter showed an oxygen
saturation level of 2.2 ± 2.3% above that of the
Ohmeda.
• The highest value measured by Nellcor was 99%
on 5,536 occasions, whereas the Ohmeda
measured a maximum value of 98% on only 40
occasions.
• Paediatric
patients.
• Comparison
between pulse
oximeters and not
with blood gases
or co-oximetry
(gold standard).
55
Authors
Study design
Outcomes
Results
Limitations
Source of funding
Trivedi et al
1997 [10]
• Population: 8 volunteers.
• Effect of ambient
light, motion and
low perfusion on
pulse oximeter
accuracy.
• The Ohmeda and Nellcor N-200 had the highest
total failure rates with respect to SpO2 and HR
due to ambient light interference. The Nellcor
Clock, Novametrix and Criticare performed
approximately equally well with regard to SpO2
determinations, however, Nellcor C Lock was the
best with regard to HR measurements.
Not specified.
• Aim: To evaluate 5 of the most commonly used
models of pulse oximeters and compared their
accuracy during low perfusion states, motion of
oximeter probe and interference by ambient light.
• Method:
-
Each volunteer was continuously monitored with
an ECG, automatic noninvasive blood press cuff
and radial arterial cather, which was placed to
measure blood pressure.
-
The 5 pulse oximeters were simultaneously
attached to each subject. The oximeters used
were Nellcor N-200, Nellcor with C Lock,
Ohmeda 3740, Novametrix Oxypleth and
Criticare 504-US.
-
Arterial blood samples were drawn
simultaneously and a co-oximeter was used as
a gold standard for measurement of SaO2.
-
The effect of ambient light was assessed by
shining a standard OT light on the pulse
oximeter probes from a distance of 4 feet. Data
was recorded for a period of 120 seconds for
each subject. A SpO2 value greater than 4%
from SaO2 value was defined as an error. HR
errors were also determined, with error defined
as a difference of 5%.
-
The effect of motion was studied using a motion
generator to move hand. SpO2 and HR errors
were recorded.
-
The effect of low perfusion was studied by
delivering a controlled pressure to each
subject’s brachial artery. Pressure was adjusted
until the mean arterial pressure measure 50-60
mmHg. SpO2 and HR errors were recorded.
• Nellcor N-200 and C Lock were the most
accurate with regard to SpO2 during 2 Hz and 4
Hz motion. The Ohmeda and Criticare oximeters
were most accurate in measuring HR during 2
Hz motion, however all oximeters failed at 4 Hz
motion.
• During states of low perfusion, the Nellcor N-200
and C Lock performed the best in terms of SpO2
and HR. The Ohmeda and Novametrix failed
approximately one-third to one-half of the time,
while the Criticare oximeter failed 100% of the
time.
56
Authors
Study design
Outcomes
Results
Limitations
Source of funding
Wood et al
1997 [11]
• Population: 20 men, 11 of whom were well-trained
elite endurance cyclists.
• Correlation of
SaO2 and SpO2
under different
levels of
exercise.
• At rest, the correlation between the SpO2 and
SaO2 was weak for both pulse oximeters, but at
the 7th minute of exercise and at VO2max the
correlations were significant.
Not specified.
• Aim: to compare the performance of Ohmeda Biox
3700e and Criticare 504 USP pulse oximeters with
arterial blood sampled during maximal cycling
exercise.
• Method:
-
Each subject performed two cycle ergometer
VO2max tests in a hypobaric chamber. The tests
were conducted with a minimum of 24 hour rest
interval between tests and all subjects were
tested within 5 days.
-
The elite cyclists began cycling at 200 watts
whilst non-elite cyclists began at 100 watts;
both increased by 25 watts each minute until
volitional exhaustion. The hypobaric chambers
were set at 745 mmHg or 695 mmHg.
-
Before each test a catheter was inserted into
the brachial artery. Samples were taken and
immediately placed on ice and analysed within
20 minutes.
-
ECG was monitored via 3-lead ECG on
Criticare monitor.
-
Arterial blood samples and oximetry recordings
were taken simultaneously at rest, during final
30 seconds of the 7th minute of exercise and
during the last 30 seconds of the workload
corresponding to VO2max.
-
The clarity of the pulse waveform on each
pulse oximeter was continually monitored to
avoid effects of motion artefact and decreased
perfusion.
-
HR was also monitored via a Polar Electro
PE3000 Sportester.
-
Any SpO2 measurement in which the HR from
pulse oximeter did not correspond to Sportester
HR ±10% was discarded from analysis.
• The bias (<3%) was relatively unchanged from
rest to maximum exercise of both oximeters,
while precision was worse during maximum
exercise than during rest.
• The highest correlations between co-oximetry
and pulse oximetry vales was found when data
from rest and exercise were combined, with
Criticare oximeter have an higher correlation
than the Ohmeda. The Ohmeda underestimated
SaO2 at all saturation levels, with the differences
greater at lower saturation levels, while the
Criticare slightly overestimated SaO2 at all levels.
• Both oximeters provided consistent estimates of
oxygen saturation, though the Criticare oximeter
was the better of the two models, with smaller
bias and superior precision.
• Male population.
57
Authors
Study design
Outcomes
Results
Limitations
Source of funding
Chiappini et
al 1998[12]
• Population: 100 subjects.
• Accuracy of
SpO2
measurements.
• The co-oximeter measured higher values of
SaO2 in comparison with N-20P oximeter, with a
statistically significant difference when using the
paired t-test.
• Male dominated
population.
Not specified.
Not specified.
• Aim: To assess the accuracy of the Nellcor N-20P.
• Method:
RheineckLeyssius
and
Kalkman
1999 [26]
-
Arterial blood samples were collected according
to the criteria of the British Thoracic Society.
-
Each sample was immediately any using a
Radiometer ABL-330 and a IL-282 Co-oximeter.
-
During the collection of the blood sample, SpO2
was measured with the pulse oximeter.
• Population: 53 patients.
• Aim: To investigate the incidence of false alarms in
the operator theatre (OR) generated by the Nellcor
N-3000 compared with a conventional pulse
oximeter (Criticare 504).
• Method:
-
In each patient was simultaneously monitored
with three pulse oximeters in the OR: Nellcor N3000 and two Criticare 504 oximeters with
different signal averaging times (21 seconds
and 3 seconds).
-
SpO2 alarms limits were set at 90%.
-
The anaesthesiologist recorded manually
episodes of probable hypoxaemia together with
the time of occurrence.
-
Criteria for detecting hypoxaemia was: cyanosis;
alarm of at least one pulse oximeter and
decreasing SpO2 values of at least one of the
two other pulse oximeters.
-
Alarm was considered to be caused by artefact
and not hypoxaemia if the pulse oximeter was
unable to determine the appropriate pulse when
compared with ECG HR.
• The difference between the co-oximeter and
pulse oximeter measurement was 1.56% and the
limits of agreement ranged from 1.08 to 4.20.
• Lack of accuracy of N-20P oximeter in
comparison with a standard reference method is
appreciable only for values at the extremes of
the oxyhaemoglobin saturation range.
• Number of
alarms.
• Cause of trigger.
• The Nellcor and the Criticare with 21 second
averaging both generated one false alarm
(duration of 55 seconds and 5 seconds,
respectively).
• In 8 patients, the Criticare oximeter with 3
second averaging produced 20 false alarms
(duration ranged from 5 to 95 seconds).
• Nellcor generated loss of pulse alarm in 3
patients, whilst the Criticare oximeter with 21
second averaging generated loss of pulse alarm
in 4 patients and with 3 second averaging in 5
patients.
• 36 of the 53 patients experienced one or several
episodes of hypoxaemia. All hypoxaemic
episodes triggered the alarm of at least one
monitor.
• Hypoxaemia triggered the alarm of the Criticare
oximeter with 3 second averaging more often
than the alarms of the other two monitors.
• 50 hypoxaemic events triggered the alarm of all
three oximeters, 19 events two of the oximeters
and 31 events just one of the oximeters.
• Comparison
between pulse
oximeters and not
with blood gases
or co-oximetry
(gold standard).
58
Authors
Study design
Outcomes
Results
Limitations
Source of funding
Bohnhorst et
al 2000 [27]
• Population: 17 spontaneously breathing unsedated
preterm infants.
• Missed alarms of
new technology
for detection of
hypoxaemia and
bradycardia.
• 202 falls in PTcO2 to <40 torr occurred, with a
range of 1 to 34 episodes per infant. Of these,
174 were identified by all 3 oximeters. In 15 of
these episodes the N-200 alarmed because of
signal loss. This happened 4 times with the
OXISMART and only once with Masimo SET.
Not specified.
• Aim: To determine whether the false alarm rates of
the Masimo SET and Nellcor OXISMART might
result in an increased proportion of missed true
alarms such as an impaired detection of hypoxaemia
or bradycardia.
• Method:
-
3 pulse oximeters were used: conventional
pulse oximeter (Nellcor N-200), oximeter with
OXISMART technology (Nellcor N-3000) and
oximeter with SET technology (Masimo SET).
-
All sensors were shielded against ambient light
and were resited in 4 hour intervals.
-
Postductal PTcO2 was measured using a
standard PTcO2 monitor (Kontron 7640).
-
HR was recorded using a standard ECG monitor
(Kontron 7271).
-
The SpO2 and PR readings from the oximeters
were recorded together with the data from the
PTcO2 and ECG monitor using purpose-written
software. The raw red and infrared adsorption
signal from the Masimo was also recorded.
-
Recording were analysed for all episodes during
which PTcO2 fell to <40 torr. Episodes were then
analysed to see if a fall in SpO2 <85% occurred
within 2 minutes of fall in PTcO2 to <40 torr.
-
Recordings were also analysed for the
bradycardias that were not accompanied by a
fall in PTcO2 to <40 torr.
• Of the 28 episodes where an oximeter did not
alarm, manual analysis of red-to-infrared ratios
showed that SpO2 had been <85% for <10
seconds or had not been <85% at all in 16
episodes.
• 11 out of 185 true hypoxaemic episodes were
missed by at least one oximeter. All 11 were
identified by the conventional oximeter, whereas
10 were missed by Nellcor OXISMART and 1
was missed by the Masimo SET.
• The mean interval between PTcO2 falling to <40
torr and SpO2 falling to <85% was -0.7 seconds
for conventional oximeter and -0.4 seconds for
both new generation devices. On average all 3
oximeters alarmed some 0.4-0.7 seconds before
the PTcO2 monitor reached the corresponding
alarm threshold.
• There were 54 bradycardias in 11 infants that
were not accompanied by hypoxaemia. Only 14
of these were identified by all 3 oximeters. 37
were missed by the Nellcor OXISMART, 4 by
Masimo SET (all had also been missed by
OXISMART) and 17 by N-200 (14 of these had
been missed by OXISMART).
• Paediatric
patients.
• Comparison
between pulse
oximeters and not
with blood gases
or co-oximetry
(gold standard).
59
Authors
Malviya et al
2000 [28]
Study design
Outcomes
Results
Limitations
Source of funding
• Population: 75 children.
• Frequency and
duration of
events.
• The overall alarm frequency was once every
36minutes for N-200 and once every 30 minutes
for the Masimo SET.
• Paediatric
patients.
Not specified.
• Aim: To compare the incidence and duration of false
alarms between Masimo SET and conventional
pulse oximeter in young children in the
postanesthesia care unit (PACU).
• Method:
• There were 27 true alarms, which were all
identified by the Masimo SET. Only 16 of the
true alarms were identified by the N-200.
-
Two probes were placed on hand of each child
and connect to Nellcor N-200 (conventional
pulse oximeter) and Masimo SET pulse
oximeter. SpO2 and HR were recorded
continuously from these devices onto bedside
computer.
• Spurious desaturations and pulse rate changes
triggered false alarms most commonly occurred
with significant gross movement. There were
more than twice the number of false alarms
exhibited by the conventional pulse oximeter
compared with the Masimo SET.
-
Additional ECG HR was recorded continuously
and simultaneously via SpaceLabs monitors.
-
Nurses were blinded to Masimo SET
measurements and made all intervention
decision and clinical observation on N-200.
• There were 52 data dropout events for the
Masimo SET compared with 44 for the N-200.
This difference was not deemed significant.
-
A research assistant concurrently documented
events such as removal of endotracheal tube or
other airway device, patient motion, nursing
interventions including placement of an artificial
airway or oxygen administration, and blood
pressure monitoring that could interfere with
signal transmission.
-
All data was coded and analysed for frequency
and duration of event: data dropout – complete
interruption of continuous SpO2; false negative
– true alarm missed; false alarm – decrease in
SpO2 <90% that by expert clinical observation
is considered incorrect; true alarm.
• Comparison
between pulse
oximeters and not
with blood gases
or co-oximetry
(gold standard).
60
Authors
Study design
Outcomes
Results
Limitations
Source of funding
Barker 2002
[29]
• Bias and
precision of
SpO2
measurements.
• There was a wide range of oximeter performance
with PI values varying from 94% to 28%. The top
5 devices in PI were the Masimo SET (94%), the
Philips Viridia 24C (84%), the Philips CMS
(80%), the Datex-Ohmeda 3740 (80%) and the
Datex-Ohmeda 3800 (79%).
• Comparison
between pulse
oximeters and not
with blood gases
or co-oximetry
(gold standard).
Masimo Corporation.
• Aim: To compare the performance of recent version
of motion-resistant oximeters during motion and
hypoxaemia.
• Method:
-
-
-
Each subject was monitored with 6 oximeter
sensors: 3 on the moving test hand and 3 of the
same make and model on the stationary control
hand.
20 different devices were tested: Masimo SET
(V2), Philips HP Viridia 24C (Rev B), Philips HP
CMS (Rev B), Datex-Ohmeda 3740, DatexOhmeda 3800, Datex-Ohmeda AS/3, Nellcor N395 (v1620), Datex-Ohmeda 3900, Novametrix
MARS (2000-10), Hewlett-Packard CMS,
Nellcor N-180, Marquette 8000, Nellcor NPB295, Novametrix 520A, Nellcor N-200, BCI
3304, Nonin 8600, SpaceLabs 90308, Nellcor
NPB-190 and Criticare 5040.
The test hand was strapped to a motorised
motion table, which produced repeatable,
continuous hand motions. It could be configures
so that the fingertips either tapped or rubbed on
a smooth surface. The amplitude of the motion
was ±2 cm and the frequency was either fixed or
randomly varied within the range of 1 to 4 Hz.
-
After recording room air control values with both
hands stationary, the motion table was activated
and 2 min of data were recorded for each of the
2 motions.
-
Hypoxaemic episodes were measured during
each motion by having the subject breath via an
anaesthesia machine, where the amount of
oxygen inspired could be controlled.
• Dropout rate.
• Performance
index (PI).
• Sensitivity and
specificity for the
detection of
hypoxaemia.
• Sensitivity to hypoxaemia ranged from 98%
(Masimo) to 28% (BCI 3304) and specificity
ranged from 93% (Masimo) to 15% (Criticare
5040).
• The devices with lower sensitivity and specificity
values had small dropout rate, meaning they
continued to display a SpO2 value that was
grossly in error during motion.
• The Masimo SET oximeter yielded the highest
values of all the calculated performance
statistics.
• More recent pulse oximeters outperformed the
older models in terms of both accuracy and
reliability during motion.
• Nellcor N-3000 was found to miss 5.4% of
hypoxaemia episodes and 69% of bradycardia
episodes, whereas the Masimo SET missed
0.5% and 7%, respectively.
61
Authors
Study design
Outcomes
Results
Limitations
Source of funding
Bohnhorst et
al 2002 [13]
• Population: 56 term and preterm infants who had
been admitted to the neonatal intensive care unit.
• Sensitivity and
specificity for
detection of
hyperoxaemia.
• 280 SpO2/SaO2/PaO2 determinations were
performed for Agilent Viridia and 291 each for the
Masimo SET and Nellcor N-3000.
Not specified.
• Aim: To determine the sensitivity and specificity of
three recently available pulse oximeters in the
detection of hyperoxaemia.
• Method:
-
The three devices studied were the Agilent
Viridia M3, Masimo SET and Nellcor N-3000
Oximsmart.
-
Care was taken to ensure sensors had good
contact with the skin and were shielded against
each other and ambient light.
-
Whenever an arterial blood sample was taken
for clinical purposes, the SpO2 readings on the
oximeters were recorded at the precise moment
the blood was drawn.
-
Measurements were not timed specifically to
include high PaO2 values or exclude periods
with motion and/or differences between HR and
PR. But SpO2 readings had to be stable for at
least 20 seconds before arterial blood was
drawn.
-
Blood samples were analysed of PaO2 using
blood gas analyser and SaO2 using co-oximeter.
-
Sensitivity was calculated as the proportion of
instances with a PaO2 >80 mmHg that were
associated with an SpO2 above the threshold,
divided by all instances with PaO2 >80 mmHg.
-
Specificity was calculated as the proportion of
PaO2 readings ≤80 mmHg associated with an
SpO2 value below the threshold, divided by the
number of instances PaO2 ≤80 mmHg.
• Accuracy and
precision of
SpO2.
• 105 in 27 patients showed a PaO2 >80 mmHg.
• At an upper alarm limit of 95%, all three devices
detected 93-95% of hyperoxaemic episodes
(sensitivity). Specificity at this threshold values was
more variable, ranging from 26% (Nellcor) to 46%
(Masimo).
• The two highest PaO2 values that were >80 mmHg,
but associated with an SpO2 ≤95% (false negative)
were 144 and 92 mmHg for Agilent, 169 and 98
mmHg for Masimo, and 141 and 95 mmHg for
Nellcor. The lowest PaO2 with SpO2 >95% (false
positive) was 46 mmHg for Agilent, 56 mmHg for
Masimo and 50 mmHg for Nellcor.
• Measurement precision was similar between the
devices, while the bias was smaller with the Masimo
and Agilent than with the Nellcor pulse oximeter.
62
Authors
Study design
Outcomes
Results
Limitations
Source of funding
Durbin and
Rostow
2002 [15]
• Population: 13 postoperative adults patients in
whom conventional pulse oximetry failed.
• Whether Masimo
SET could
capture SpO2
when
conventional
pulse oximetry
was unable too.
• Masimo SET was able to obtain pulse oximetry
readings in 12 of the 13 patients.
Not specified.
• Aim: To test the ability of the Masimo SET
technology to acquire and maintain reliable pulse
oximetry signals in critically ill, postoperative patients
in whom conventional pulse oximetry technology
was unable to provide reliable monitoring.
• Method:
-
Failure of conventional pulse oximetry was
defined as complete inability of oximeter to
obtain a pulse signal, or display of an obviously
spurious saturation or inaccurate pulse rate
value.
-
Immediately following failure, an oximeter
incorporating Masimo SET (IVY 2000) was used
in attempt to acquire a pulse oximetry signal.
-
When and if a stable pulse oximetry reading
was obtained, an arterial blood gas sample was
obtained for evaluation and validation of the
SpO2 and the pulse rate was confirmed by
comparison with ECG HR.
• Accuracy of
Masimo SET
SpO2 and PR
measurements.
• The SaO2 to SpO2 difference was 1.1% ± 1.0%
(mean ± SD) for these patients.
• Two patients exhibited SpO2 values that were
inaccurately high. The specific cause of errors
was unclear.
• In the one patient in whom a reliable reading with
Masimo SET was not obtained, blood gas data
could also not be obtained as the patient went
into cardiac arrest.
63
Authors
Study design
Outcomes
Results
Limitations
Source of funding
Durbin and
Rostow
2002 [14]
• Population: 86 patients undergoing coronary artery
revascularisation surgery with good preoperative
ventricular function.
• Patient care
outcomes.
• A total of 82 patients were available for analysis
of clinical management: 43 had Masimo SET
pulse oximetry data displayed and 39 has
conventional pulse oximetry data displayed.
• Data was not
available for all
the patients
enrolled in the
study.
No specified.
• Aim: To investigate whether improved oximetry
would result in a measurable difference in patient
care outcome variables compared with conventional
less reliable oximetry.
• Method:
-
Patients were monitored continuously with a
Masimo SET pulse oximeter (innovative
technology) and Ohmeda 3740 (conventional
pulse oximeter). Data was recorded on a laptop.
-
Patients were randomised to have the output
from one of the devices displayed for use by the
bedside caregiver.
-
Pulse oximetry failure percentage was used as
a measure of the efficiency of monitoring.
Failure percentage was defined as percentage
of total monitoring time when the monitor output
was unreliable (no signal, obvious artefact or
difference of >10% between SpO2 and SaO2
(calculated from PaO2)).
-
Arterial blood gas determinations were used to
calculate the accuracy and precision of
oximeters.
• Reliability of
oximeters.
• There was no difference in the time to extubation
between the two groups, however, patients in the
Masimo SET group were weaned to a FIO2 of
0.4 an hour sooner and with significantly fewer
arterial blood gas analyses.
• Only 59 patients had adequate data for reliability
evaluation. The conventional oximeter
experienced almost 8 times more failure time
than Masimo.
• Comparison of Masimo with blood gas gave a
bias of 0.53 ± 1.7 and for conventional oximeter 0.82 ± 2.8.
• Accuracy
determined by
comparison with
PaO2 rather than
SaO2.
64
Authors
Study design
Outcomes
Results
Limitations
Source of funding
Gehring et al
2002 [30]
• Effect of motion
and low
perfusion.
• For desaturation with no motion and normal
perfusion the differences between the oximeters
were within ±3%.
Agilent
Technologies/Philips
Medical Systems,
Datex-Ohmeda and
Nellcor/Tyco
Healthcare Group.
• Aim: To determine if new-generation oximeters
improve the accuracy of tracking SpO2 and PR
during induced patient hypoxaemia in the presence
of motion artefact and low perfusion.
• Method:
-
The oximeters tested included: Datex Ohmeda
3900P, Agilent CMS, Nellcor N-395, Ivy
SatGuard 2000 with Masimo SET and Nellcor
N-3000.
-
The reference oxygen saturation was recorded
on the right hand using 2 Nellcor N-3000 pulse
oximeters. The left hand was subjected to
motion and low perfusion. HR was recorded
with an ECG, using electrodes place on the
chest.
-
Participants breathed through a tight-fitting face
mask connected to a Trajan 808 with fresh gas
supply. Fraction of inspired oxygen was varied
by adjusting oxygen and nitrogen. Inspiratory
and expiratory oxygen and carbon dioxide
concentrations and breathing patterns were
continuously monitored.
-
Motion generator was used to move test hand
in irregular patterns. Volunteers were also
asked to perform a tapping motion for 120
seconds, followed 60 seconds later by a
scratching motion for 120 seconds.
-
Low perfusion was induced by compression of
the brachial artery in the upper arm.
-
The subject’s oxygen saturation started at
100% and was decreased to 90%, where it was
held for 1 minute. It was then decrease to 80%.
It was then increased back to 100%.
-
SpO2, PR and ECG HR were continuously
recorded.
• The SpO2 data from the various oximeters of the
test and reference hand were in good agreement
as perfusion changed.
• In the period with motion and low perfusion, all
oximeters tested showed high SpO2 errors.
• The maximum positive deviation between true
SpO2 and measured SpO2 was 30% at the
lowest saturation tested.
• Datex-Ohmeda device exhibited the worst pulse
indication performance followed by the Nellcor
N-3000. There were no significant differences
between the Agilent, Ivy and N-395.
• Pulse oximeters reliably measured SpO2 when
movement was introduced as an artefact and
when perfusion was reduced. Only during the
clinically extreme condition of combined low
perfusion and motion was the reliable detection
of SpO2 <63%. During combined reduced
perfusion and motion, 3 of the oximeters proved
to reliably show PR, whereas N-3000 and DatexOhmeda 3900P showed clear differences from
ECG HR.
• Comparison
between pulse
oximeters and not
with blood gases
or co-oximetry
(gold standard).
65
Authors
Study design
Outcomes
Results
Limitations
Source of funding
Hay et al
2002 [31]
• Population: 26 neonates in initial study and 7
neonates in the follow-up study.
• Monitoring
performance of
Masimo SET
compared with
Nellcor N-200.
• Both Masimo SET and Nellcor N-200 pulse
oximeters produced similar SpO2 and PR values
during nonmotion conditions. However, only the
Masimo SET instruemtn recorded a stable value
of SpO2 and captured a PR similar to the ECG
HR during motion conditions.
National Institutes of
Health GCRC Grant.
• Aim: To evaluate the effects of motion on the
incidence of SpO2 and PR alarms in a neonatal
population, comparing the results obtained from a
conventional pulse oximeter (Nellcor N-200) to those
of a Masimo SET pulse oximeter.
• Method:
-
The Masimo sensor, LNOP Neo, and the Nellcor
sensor, Oxisensor II N-25, were attached to
different feet of each infant. To reduce effect of
site bias, the sensors were switched to the
opposite foot after 3 hours and monitored for an
additional 3 hours.
-
The study was blinded to provide an objective
comparison of pulse oximeter performance
without affected routine care of the infant.
-
The ECG HR from a Datascope Passport
monitor was collected to corroborate each pulse
oximeter’s PR.
-
Potential sources of artefact, including infant
motion, nursing care and intervention, and
parental handling and care, were logged by the
research nurse.
-
Data were reviewed without knowledge of which
pulse oximeter was used to determine how well
each met the clinical indices for monitoring
neonates, specifically, true and false
desaturation, and bradycardic events.
-
Following initial study, the performance of three
new-generation pulse oximeter (Nellcor N-395,
Novametrix MARS and Philips Viridia 24C) was
compared to Masimo SET.
-
The hypoxaemia alarm threshold was set at
85%. Bradycardia was defined as an ECG HR
of ≤80 bpm
• Comparison of
performance of 4
new-generation
pulse oximeters.
• R/IR analysis found that the Masimo SET
provided accurate measures of SpO2 and PR.
• The Nellcor N-200 had a total of 396.6 minutes
of zero outs and false alarms (4.2% of operation)
across all the infants. The Masimo SET pulse
oximeter exhibited 31.3 minutes of zero outs and
false alarms (0.3% of operation).
• The Masimo SET had significantly fewer false
alarms than the Nellcor and they were shorter in
duration, resulting in 92% less total alarm time.
• 8 infants had a total of 14 transient bradycardic
episodes. Of these, the Masimo SET pulse
oximeter caught significantly more, 12 compared
to 2 for the Nellcor.
• In the comparison of the f new-generation pulse
oximeters, false desaturations totalled 86 (1, 10,
33 and 42 for Masimo, Nellcor, Philips and
Novametrix, respectively), while missed
desaturations were few (1, 4, 6 and 12 for
Masimo, Nellcor, Philips and Novametrix,
respectively). These new-generation devices
differed greatly in their ability to detect changes
in PR (ie the frequency of frozen PR during times
of ECG HR change was 0, 6, 11 and 46 for
Masimo, Nellcor, Philips and Novametrix,
respectively).
• Paediatric
patients.
• Comparison
between pulse
oximeters and not
with blood gases
or co-oximetry
(gold standard).
66
Authors
Study design
Outcomes
Results
Limitations
Source of funding
Irita et al
2003 [32]
• Population: 18 patients.
• Pulse oximeter
failure.
• Pulse oximeter failure developed in 14 patients
with the AY-900P oximeter, whereas it
developed in only 4 patients with Masimo. These
4 patients also had pulse oximeter failure for the
AY-900P.
Not specified.
• Pulse oximeter failure occurred immediately after
the initiation of bypass 4 patients and just after
aortic cross-clamping in 9 patients.
• Comparison
between pulse
oximeters and not
with blood gases
or co-oximetry
(gold standard).
• Aim: To compare the performance of a newgeneration pulse oximeter with that of a conventional
device during cardiopulmonary bypass under
moderate hypothermia.
• Method:
-
Nihon Kohden AY-900P was the conventional
pulse oximeter used and the new-generation
pulse oximeter was the Masimo SET Radical.
-
Pulse oximeter data was collected in real-time
with a PC data acquisition system and
handwritten notes.
-
Pulse oximeter failure was defined as failure to
show no SpO2 data and/or to display incorrect
SpO2 for longer than 3 minutes continuously.
Incorrect SpO2 was defined as <97% during
bypass, because arterial saturation obtained
with co-oximetry was >99%. The duration of
pulse oximeter failure was calculated as the
duration of no SpO2 plus the duration of
incorrect SpO2.
-
The dropout rate corresponded to the
percentage of time when the pulse oximeter
shed no SpO2 data.
-
Pulse oximeter performance was also examined
from the standpoint of preoperative diuretic
therapy and intraoperative hyperlactatemia.
• The duration of pulse oximeter failure was 36% ±
31% of the duration of bypass for AY-900P
oximeter and 6% ± 15% for Masimo. The
duration of pulse oximeter failure was 46% ±
43% of duration of aortic cross-clamping for AY900P and 5% ± 15% for Masimo.
• No SpO2 was provided for 36% ± 39% of the
duration of aortic cross-clamping for AY-900P
and 4% ± 12% for Masimo. The observed pulse
oximeter failure was mainly due to no SpO2
rather than incorrect SpO2.
• Of 6 patients with preoperative diuretic therapy, 5
developed pulse oximeter failure in AY-900P, but
only one in Masimo.
• The incidence of pulse oximeter failure was
similar between the AY-900P and Masimo
among 12 patients without preoperative diuretic
therapy.
• Of the 9 patients with lactate less than 80 mg/dL,
pulse oximeter failure developed in 3 patients
with AY-900P and 2 with Masimo.
• Of the 9 patients with lactate greater than 90
mg/dL, pulse oximeter failure developed in 8
patients with AY-900P and 2 with Masimo.
• Did not consider
accuracy or
precision of SpO2
measurements.
67
Authors
Study design
Outcomes
Results
Limitations
Source of funding
Sahni et al
2003 [33]
• Population: 15 healthy male infants.
• Effect of motion
of SpO2 and HR.
• The incidence of artefact was significantly lower
with the Masimo than with the Nellcor during and
after circumcision. Even during baseline period
when the infants were quietly asleep, there was
a threefold higher incidence of artefact with the
Nellcor.
Not specified.
• Mean SpO2 and HR values obtained for the
Masimo were significantly higher than those
obtained with the Nellcor, especially during
circumcision, when extreme motion was likely.
Similar differences were observed after
circumcision.
• Comparison
between pulse
oximeters and not
with blood gases
or co-oximetry
(gold standard).
• Aim: To compare the effects of motion on
measurements of SpO2 and HR made with Masimo
SET, with the same measurements made with the
Nellcor N-200.
• Method:
-
Continuous pulse oximetry and HR monitoring
were performed using Masimo SET and Nellcor
N-200 pulse oximeters and a standard HR
monitor (Hewlett-Packard 3680).
-
Baseline data were collected for 10 minutes with
the infant quietly asleep. Data collection was
then continued during and after circumcision for
a total duration of one hour.
-
Simultaneous minute by minute assessments of
behavioural sleep and activity state were also
made.
• The Masimo HR signal predicted the ECG HR
more accurately, with lower residual error.
• The HR and SpO2 measured by Nellcor were
lower and more variable during all behaviour
states. The effects were more pronounced
during behavioural activity states associated with
excessive motion artefact.
• Higher HR values were recorded by the Masimo
during non-sleeping states; however, the Nellcor
recorded not only much lower but more variable
HR values during active sleep, awake and crying
states than those recorded during quiet sleep.
This was also reflected in the SpO2
measurements, which remained stable for
Masimo but were significantly lower for Nellcor
during all behavioural activity states.
• Paediatric
patients.
• Only male
patients.
68
Authors
Study design
Outcomes
Results
Limitations
Source of funding
Robertson
and
Hoffman
2004 [34]
• Population: 27 patients (age range 2 days to 15
years) monitored in neonatal intensive care,
paediatric intensive care unit, operating theatre and
sleep laboratory.
• Time of data
availability.
• 308,301 s of data acquired, with data reporting
highest for Solar 8000 displaying data 98.7% of
overall time, compared with 98.4% for the
Masimo SET and Nellcor N-395.
Not specified.
• Aim: To compare data availability, signal heuristics,
and agreement of two new oximeter technologies
compares with an older standard.
• Signal heuristics
and warnings
stratified by both
signal integrity
and SpO2.
• Method:
-
A blinded side-by-side comparison of Masimo
SET, Nellcor N-395 and GE Marquette Solar
8000 oximeters.
-
The Masimo LNOP Neo sensor was used with
the Masimo oximeter and Nellcor Oxisensor N25 sensors were used with the N-395 and Solar
8000 oximeters.
-
Study tested the motion artefact, loss of pulse
and signal quality indicators of the Masimo and
N-395 oximeters.
-
Data from all oximeters were compared.
-
Co-oximetry data were not collected during the
study.
• Measures of
agreement.
• Masimo reported data 97.1% of time overall with
no signal heuristic flag. Nellcor N-395 reported
data 84.7% of overall time with no signal
heuristic flag.
• Agreement between the devices deteriorated at
a SpO2 of less than 80% and grew progressively
worse at lower saturations.
• Under optimal signal conditions, with no signal
heuristic displayed, there was little difference in
precision and bias between the two newer
technologies, but agreement between devices
was significantly affected by signal quality,
motion artefact and hypoxaemia.
• Masimo device posted less questionable data
that the Nellcor N-395 device.
• The Solar 8000 detected all hypoxaemia
episodes, whereas 5.4% were missed by the
Nellcor N-395 and 0.5% by the Masimo device.
• In the sleep laboratory, of 75 desaturation events
not related to movement artefact, 98.6% of
events were detected by Masimo oximeter,
whereas Nellcor N-395 detected 45.3%.
• Paediatric
patients.
• Comparison
between pulse
oximeters and not
with blood gases
or co-oximetry
(gold standard).
69
Authors
Study design
Outcomes
Results
Limitations
Source of funding
Torres et al
2004 [16]
• Population: 46 children: 25 children with congenital
heart disease undergoing surgical repair or palliation
using cardiopulmonary bypass; 21 children with
cyanotic congenital heart disease.
• Bias and
precision of
pulse oximeters.
• In the postcardiopulmonary bypass group,
Nellcor N-395 failed to measure SpO2 during
41% of the SaO2 measurements compared to
10% for the Masimo SET Radical.
Department of
Pediatrics,
University of Illinois
College of Medicine.
• Aim: To determine whether SpO2 measured with the
Nellcor N-395 and the Masimo SET Radical pulse
oximeters were accurate and/or precise when
compared with SaO2 of the whole blood of children
immediately recovering from cardiopulmonary
bypass, a transient period of poor peripheral
perfusion, or in children with persistent hypoxaemia
(ie SaO2 <90%).
• Method:
-
SpO2 displayed on each pulse oximeter at the
time routine arterial blood SaO2 measurements
were recorded. SaO2 was measured using a cooximeter.
-
Cardiopulmonary monitor HR, pulse oximeter
HR and failure of pulse oximeter to display SpO2
at the time of blood gases measurement were
also recorded.
• There was a significant difference in bias ±
precision (Nellcor 1.1 ± 3.3; Masimo -0.2 ± 4.1)
but no significant difference in HR between
oximeters (Nellcor 0.7 ± 6.9; Masimo -0.6 ± 1.9).
• 17 of the 61 SaO2 measurements in the bypass
group were less than 90%. There was a
significant difference in bias and precision when
compare SaO2 <90% versus SaO2 ≥90% for the
Nellcor N-395 and Masimo SET Radical.
• In the cyanotic congenital heart disease group,
there were no SpO2 failures for either oximeter.
There was no significant difference in bias ±
precision between oximeters (Nellcor 2.9 ±4.6;
Masimo 2.8 ± 6.2). The mean HR differences
were not significantly different (Nellcor -0.1 ± 2.0;
Masimo -0.05 ± 1.8).
• There was a loss of accuracy and increase in
scatter (increase of bias and precision) for both
oximeters when SaO2 <90%.
• Masimo SET Radical and Nellcor N-395 pulse
oximeters have an acceptable accuracy but
unacceptable precision in children with
congenital heart disease.
• Paediatric
patients.
70
Authors
Study design
Outcomes
Results
Limitations
Source of funding
Bickler et al
2005 [17]
• Population: 21 healthy, non-smoking subjects: 11
with dark pigmentation of skin and 10 light
pigmentation.
• Affect of skin
pigmentation.
• The mean of all 1,067 data points at all SaO2
levels and all three types of oximeter indicated
that SpO2 read approximately 1% higher in darkthan in light-skinned subjects. The overall mean
bias errors due to pigment were +0.4% with
Nonin, +0.6 with Novametrix and +1.6% with
Nellcor.
UCSF Department
of Anesthesia
research fund.
• Aim: To determine the affect of skin pigmentation on
pulse oximeter accuracy.
• Method:
-
Subjects were studied semisupine with a nose
clip while deliberately hyperventilating airnitrogen-CO2 mixtures via a mouthpiece from a
partial rebreathing circuit.
-
A radial artery catheter was placed to facilitate
arterial blood sampling for measurement of
SaO2.
-
5 oximeters were mounted on each subject: one
Nellcor N-595, two Novametrix 513s and two
Nonin Onyxs. Measurements from the
Novametrix and Nonin oximeters were recorded
manually.
-
The Nellcor oximeter was tested in all subjects
and the Novametrix and Nonin oximeters were
tested in 9 light- and 7 dark-skinned subjects.
-
A series of 10-12 stable target SaO2 plateaus
between 60 and 100% were sought by an
operator adjusting the inspired air-nitrogen-CO2
mixture breath by breath in response to an
analogue meter displaying the estimated SaO2.
-
At each level, arterial blood was sampled after a
plateau of 30-60 seconds, followed by a second
sample at the same plateau 30 seconds later.
• At lower oxyhaemoglobin saturation, greater
differences in bias between light- and darkskinned subjects were seen.
• With all three devices the effect of skin
pigmentation on bias increased approximately
linearly as SaO2 decreased.
• In the range of 60-70% SaO2, the mean
difference in bias between light- and darkskinned subjects was +1.4% (Nonin), +4.4%
(Novametrix), and +4.3% (Nellcor).
• Dark skin pigment-related bias was statistically
significant with Nellcor in all SaO2 decades and
with Novametrix between 60 and 80%, but only
in 70-80% range with Nonin.
• Small but statistically significant differences were
found between oximeters. The mean bias of
Novametrix (1.40 ± 1.94%) was statistically
higher than that of Nellcor (0.49 ± 2.18%) and
Nonin (0.08 ± 1.45%). Small differences were
also found within decadal SaO2 intervals.
• Dark skin pigmentation results in overestimation
of arterial oxygen saturation, especially at low
saturation in the 3 tested oximeters.
• Study focused
more on the
cause of
inaccuracy rather
than which
oximeter was
most accurate.
71
Authors
Study design
Outcomes
Results
Limitations
Source of funding
Murthy et al
2005 [18]
• Population: 20 patients in whom clinicians could not
get a conventional pulse oximeter signal. Patients
ranged from neonates to elderly.
• Correlation
between Masimo
and arterial
blood gas
analysis.
• Values measured by Masimo correlated well with
those from blood gas analysis.
Not specified.
• Aim: To evaluate the performance of Masimo SET in
several conditions where conventional pulse
oximeters exhibit a low or no signal alarm,
comparing oxygen saturation values with that of
arterial blood gas analysis.
• Method:
-
Conventional pulse oximeter was Agilent
M3046A M4. Conventional pulse oximeter
declared as failed only after several attempts to
get a signal were made. Sensor was left in
place till end of study.
-
When a saturation reading and plethysmograph
was established on Masimo oximeter an arterial
blood sample was taken.
-
PR from conventional and masimo oximeters
were recorded and compared to ECG HR.
• Motion disturbance could be seen on
plethysmograph for Masimo but the oximetry
readings remained constant and match blood
gas analysis values.
• The PR of the Masimo corresponded to the ECG
HR in all patients, whereas the convention
oximeter showed erratic readings.
• Bias and precision
not defined from
results.
72
Authors
Study design
Outcomes
Results
Limitations
Source of funding
Workie et al
2005 [35]
• Population: 36 neonates in a neonatal intensive
care unit.
• Oxygen
saturation
measurement.
• One patient was removed from the study
because of equipment failure.
Not specified.
• Aim: To compare Philips FAST pulse oximeter with
the Masimo SET pulse oximeter to test the null
hypothesis that there is no difference in the efficacy
of these two new-generation motion-tolerant pulse
oximetry devices.
• Method:
-
The two pulse oximeters were used to
simultaneously monitor each neonate.
-
Both oximeters were programmed for a signal
averaging time of 10 seconds.
-
Readings were considered accurate only when
the waveforms were of good quality and/or the
pulse oximeter’s pulse rate (PR) correlated with
the ECG HR monitor.
-
Measurements were made every 5 minutes for a
period of 2 hours on each patient. Potential
sources of artefact such as patient motion,
nursing care and parental handling were also
noted.
-
Pulse oximeters were programmed to emit an
alarm for episodes of hypoxaemia (SpO2 ≤85%),
bradycardia (HR ≤80 bpm) and tachycardia (HR
≥200 bpm). True alarms were defined as
episodes that resulted in an alarm in the
presence of a good waveform and correlation of
pulse oximeter’s PR with ECG HR. False alarms
were episodes that occurred in the absence of a
good waveform and had poor correlation.
-
Dropouts were instances when the pulse
oximeter would emit an alarm due to an inability
to isolate the arterial pulse and provide an
oxygen saturation reading.
• Number of true
and false
alarms.
• Number of
dropouts and
their duration.
• 15 patients were ventilator dependent, 3 required
respiratory support from nasal continuous
positive airway pressure, 2 had nasal cannual, 1
had a tracheostomy collar with supplemental
oxygen and 14 were receiving room air. 17
patients were also receiving some form of
sedation.
• The SpO2 values for the Philips and Masimo
pulse oximeters were 96.4 and 95.8%,
respectively. There were no statistical difference
between the two devices across all patients and
time measurements.
• The two pulse oximeters were similar in the
number of true alarms and false alarms.
• The Philips pulse oximeter had a markedly
increased number of dropouts (247 versus 38 for
the Masimo unit), which was statistically
significant. The duration of dropouts was also
three times longer with the Philips pulse oximeter
(60.0 versus 20 minutes for the Masimo unit).
• Paediatric
patients.
• Comparison
between pulse
oximeters and not
with blood gases
or co-oximetry
(gold standard).
73
Authors
Study design
Outcomes
Results
Limitations
Source of funding
Richards et
al 2006 [36]
• Population: 36 adult patients admitted to the
progressive-care unit after cardiac surgery.
• Amount of data
dropout.
• The Nellcor had the highest percentage of data
dropout per patient and the Masimo the least.
Philips Medical
Systems.
• Aim: To compare the clinical performance of 3 newgeneration pulse oximeters with cardiac surgery
patients during their first postoperative ambulation.
• Number of false
alarms.
• The Masimo had the most false alarms and the
Nellcor the least.
• Heart rate
correlation.
• The Masimo was functional for the greatest
percentage (75.7%) of the time, the Philips
71.7% and the Nellcor 57.3%.
• Method:
-
-
The 3 SpO2 systems studied were the Philips
FAST system used with the M3 patient monitor
and M1191A adult finger sleeve sensor, the
Masimo SET V2 used with the LNOP DCI adult
clip sensor and the Nellcor N-3000 used with
the DS1000A adult clip sensor.
Each patient was monitored continuously and
simultaneously by all 3 pulse oximeters, as well
as by a centralized cardiac telemetry monitor.
-
The amount of dropout and the number of false
alarms were recorded manually during the
ambulation period.
-
Data dropout was defined as a loss of signal
initiating an alarm. A false alarm was defined as
a spurious saturation of ≤90% with either an
additional lack of correlation between the HR
reading from oximeter and reading from ECG or
a value that differed significantly from the other
2 devices when those devices were recording
SpO2 values >90%.
• There were statistically significant differences in
performance among the 3 device pairings for
both data dropout and false alarms.
• There were no differences in the mean device
performance with regard to HR accuracy across
any of the tested devices.
• 70% of the readings for the Philips and Nellcor
devices were within ±2 beats/min of the HR,
compared to 60% of the reading for the Masimo.
• Clinicians were
not blinded to the
oximeters.
• Comparison
between pulse
oximeters and not
with blood gases
or co-oximetry
(gold standard).
74
Authors
Study design
Outcomes
Results
Limitations
Source of funding
Feiner et al
2007 [19]
• Population: 36 healthy, non-smoking subjects, with
a range of skin pigment.
• Affect of skin
pigmentation,
gender and
probe type on
oximeter
accuracy.
• 17 female and 19 male subjects participated. 17
were dark, 7 intermediate and 12 light skinned.
Research fund
generated from the
clinical testing of
pulse oximeters.
• Aim: To determine the effect of an range of human
skim pigmentation on pulse oximeter accuracy at a
range of SaO2 from 70% to 100%; to determine
whether gender affects pulse oximeter accuracy; and
to determine whether probe type(adhesive
disposable versus clip-on) affects the accuracy of
SaO2 estimates/SpO2 values.
• Method:
-
Each subject’s skin was categorized as light,
dark or intermediate, and confirmed in subject
photographs by an observer blinded to
saturation data/oximeter performance.
-
Subjects were studied while reclining in a
semisupine position and deliberately
hyperventilating air-nitrogen-CO2 mixtures via a
mouthpiece from a partial breathing circuit.
-
A radial artery catheter was placed to facilitate
arterial blood sampling for measurements of
SaO2.
-
3 oximeters using an adhesive disposable probe
and a clip-type probe were mounted in each
subject’s fingers: Nellcor N-595, Masimo
Radical and Nonin 9700.
-
A series of 11 stable target SaO2 plateaus
between 60% and 100% were achieved by an
operator who adjusted the inspired air-nitrogenCO2 mixture breath by breath in response to the
estimated SaO2 derived from an
oxyhaemoglobin dissociation curve determined
for each subject.
-
Arterial blood was sampled after a plateau of
30-60 seconds. A second sample was taken at
the same plateau 30 seconds later.
• With the exception of the Masimo with adhesive
probe, all device and probe combinations
showed positive bias in intermediate- and darkskinned subjects at low SaO2. The greatest
degree of bias was found with the adhesive
probes.
• At the extreme, the Nellcor adhesive and the
Masimo clip-on read on average nearly 10 points
differently in dark-skinned subjects at 60%-70%
saturation and 7 point s at 70%-80% saturation.
• Pulse oximeter bias was significantly influences
by SaO2 range for all oximeters and probe
combinations. The pattern of bias varied
according to oximeter configurations, will all
except the Masimo with adhesive probe showing
increasing positive bias at low SaO2.
• Gender is a statistically significant determinant of
pulse oximeter bias, with the magnitude of
gender bias difference varying with
oximeter/probe type and SaO2 range. With 5 of
the 6 oximeter/probe combinations, females had
greater bias in saturation estimates over the
saturation range.
• The Nonin with clip-on probe and Masimo with
adhesive probe did not show an effect of skin
pigment, although both had a significant gender
effect.
• Pulse oximeters general overestimated SaO2 in
hypoxic subjects (SaO2 <80%). The bias was
greatest in dark-skinned subjects. Though
Masimo with adhesive probe underestimated.
• Study focused
more on the
cause of
inaccuracy rather
than which
oximeter was
most accurate.
75
Authors
Study design
Outcomes
Results
Limitations
Source of funding
Mathes et al
2008 [37]
• Population: 20 healthy, non-smoking volunteers.
• Signal quality.
• All monitors displayed normal pulse oximeter
signal at the beginning of the experiment. After
activation of neuronavigation equipment, all
monitors were at least partially affected by signal
interference. The cotton blanket improved signal
quality in all monitor except the Dash. Signal
disturbance diminished further with aluminium
foil for the Das, In vivo, Dräger and Philips pulse
oximeters.
• Considered the
effect of system of
signal quality
rather than the
accuracy of the
pulse oximeters.
Not specified.
• Aim: To investigate the effects of neuronavigation on
the performance of 6 different brands of pulse
oximeters.
• Method:
-
Each subject was monitored with the 6 pulse
oximeters whilst breathing room air. The 6
oximeters used were: GE Healthcare Tuffsat,
Philips FAST SpO2 module, GE Healthcare
Dash 3000 Pro patient monitor, Smith Medical
BCI 3303 handheld oximeter, Siemens In vivo
4500 Plus and Dräger Infinity Delta patient
monitor.
-
For each subject and monitor, HR, SpO2 and
signal quality were recorded. Signal quality was
defined as: no pulse wave present during whole
time of assessment (no signal detection);
disruption of the signal for one or more episodes
of >15 seconds (severe noise); one or more
episodes of <15 seconds of signal interference
(moderate artefact).
-
Cameras of a neurosurgical stereotactic
guidance system were positioned 100 cm above
subject’s hands and neuronavigation system
activated. Cotton blanket and aluminium foil
were used to shield probes to reduce
interference.
-
Measurements were recorded for 5 minutes with
neuronavigation on and for 5 minutes with it off.
• SpO2 readings displayed normal values for all
subjects at the beginning of the experiment. No
significant differences were observed among the
monitors.
• After activation of neuronavigation system, 5 of
the 6 monitors were unable to display SpO2 in
some of the subjects without shielding. The
Dräger monitor was able to display SpO2 in all
subjects. Shielding with a cotton blanket
improved saturation detection significantly for the
Dash, In vivo and Tuffsat, BCI and Philips
monitors. Aluminium foil improved detection in
the Dash, In vivo and Tuffsat monitor.
• When a SpO2 reading was present, no statistical
difference was noted between baseline values
and the ones obtained during neuronavigation.
• Comparison
between pulse
oximeters and not
with blood gases
or co-oximetry
(gold standard).
76
Authors
Study design
Outcomes
Results
Limitations
Source of funding
Levrat et al
2009 [20]
• Population: 25 patients admitted to ICU.
• Concordance of
SpO2 with SaO2.
• 83 measurements were recorded, including 62 in
sedated patients, 51 in patients requiring
norepinrphrine infusion, 19 in the presence of
movements hindering the taking of SpO2, 8 in
patients with proven hypoxaemia and 8 in
patients running a temperature below 36°C.
Not specified.
• Aim: To compare the concordance between SaO2
and SpO2 measured using Masimo SET or
conventional pulse oximeters in situations at risk of
producing defective signals with the conventional
technology.
• Method:
-
-
-
-
Masimo SET Radical pulse oximeter and Philips
M1020A pulse oximeter with Nellcor Oxymax
sensor were used to simultaneous measure
SpO2.
Measurements were recorded when one of the
two pulse oximeters did not display a value,
when the difference between the values
displayed by the oximeters was greater than five
saturations points or at any time a blood gas
analysis was done.
At each measurement the signal quality index,
the presence of movement, the use of sedative,
the use of norepinephrine, the patient
temperature and the concomitant values of
SaO2 and SpO2 were noted.
Four measurements were made per patient and
a minimum period of 3 hours elapsed between
measurements.
• Capacity of
pulse oximeter
to display valid
measurements
in clinical
situation.
• Display of SpO2 values was more frequent with
Masimo SET than with the conventional
technology.
• 89% of cases were considered valid with Masimo
SET and only 66% were valid for conventional
technology.
• Comparison with SaO2 found Masimo SET
estimates of arterial oxygen saturation more
accurate than the conventional technology.
• Removing nonvalid measures from the results
improved the performance of both oximeters,
though Masimo SET still proved the more
accurate estimations of arterial oxygen
saturation.
• For Masimo SET, in 95% of cases SpO2 is within
-3.1 and +3.2 saturation points of SaO2.
• For conventional technology, in 95% of cases
SpO2 is within -7.5 and +9.8 saturation points of
SaO2.
77
Table 12. Unpublished clinical trials for GE Healthcare pulse oximeters
Authors
Device/technology
Study design
Results
Limitations
Anonymous
2002 [38]
Datex-Ohmeda TruTrak+
• Population: 663 patients and healthy volunteers.
• Two arterial samples were taken from the radial artery
of non-moving had during each motion period. Subjects’
saturation stated at about 95% and went down to 73%.
The TruTrak+ closely agreed with co-oximeter
measurement during periods of motion. The traditional
pulse oximeter used was about 5% low. Masimo
displayed similar readings to TruTrak+.
• Not peer-reviewed.
• Aim: To compare the performance of the TruTrak+
technology to commercially available oximetry
technology (with and without motion capability) and
to arterial saturations measured by hemeoximetry.
• Method:
- Compared to traditional oximetry, Nellcor and
Masimo motion oximetry, non-moving reference
oximeters and Radiometer OSM-3 co-oximeters.
• When TruTrak+ was compared to traditional oximetry in
hospital, the traditional oximeter demonstrated less
accuracy with more freezing, dashing and dropouts.
- Testing took place in laboratories at DatexOhmeda and tat 5 hospitals in a wide variety of
environments including NICU, paediatrics, ICU,
OR, PACU and general wards.
• In the hospital comparisons to the Nellcor oximeter
during low perfusion conditions, the TruTrak+
demonstrated greater accuracy with less dashing and
freezing and fewer dropouts.
- Within the hospital environments, data was
collected for 30 minutes to 5 hours. 2 or 3
oximeters were placed on each subject with
simultaneous graphing of data. Pulse oximeter
readings were compared to arterial blood gases
when blood was drawn. No blood was drawn
specifically for the purpose of the study.
• In the comparison of TruTrak+ with traditional and
Masimo oximeters in laboratory clinical motion tests,
TruTrak+ demonstrated greater accuracy in the ability to
track a saturation change.
- In the laboratory, extremely low perfusion,
motion (clinical and mechanically imposed) and
a combination of motion with low perfusion were
created in volunteers. Different desaturation
episodes were created by subject’s breathing a
hypoxic mixture of nitrogen and oxygen.
• Not all comparisons were
made to the gold standard
of co-oximetry.
78
Table 12. Unpublished clinical trials for GE Healthcare pulse oximeters (continued)
Authors
Device/technology
Study design
Results
Limitations
Raley 2004
[56]
Datex-Ohmeda TruSat
• Population: 8 healthy, non-smoking subjects.
• For ultra low to low perfusion, accuracy root mean
square was 2.7, using data from 4 subjects.
• For ultra low to impaired perfusion, accuracy root mean
square was 2.5, using data from 7 subjects.
• Level of perfusion
determined by the pulse
oximeter under test,
rather than using an
independent system.
• Aim: To evaluate the SpO2 accuracy and precision
of the Datex-Ohmeda TruSat pulse oximeter with
OxyTip+ sensors over the SpO2 range of 70-100%
arterial blood oxygen saturations.
• Method:
- Comparison to functional oxygen saturation as
measured by co-oximetry.
- Each subject was cannulated with an indwelling
catheter in the left radial artery.
- The subject was moved to a cooled room (15-20
°C). The left side of the subject (side with arterial
line) was kept warm with a blanket and a heat
pad. The right side of the subject was cooled to a
low perfusion level using the room temperature
controls.
- Lower perfusion was measured by the TruSat
and cautions were taken so that the subject did
not shiver.
- Once subjects were cooled to an acceptable
level, baseline data was recorded. Each subject
was desaturated to various stable saturation
levels of approximately 5% decrements from
95% to 70% returning to 100% SaO2 by
controlling the fraction of inspired oxygen.
- SpO2 values from pulse oximeters were collected
simultaneously to blood being drawn from the
arterial line.
• For ultra low to normal perfusion, accuracy root mean
square was 1.9, using data from all 8 subjects.
• The SpO2 data collected showed that the TruSat pulse
oximeter supports a claim of 3% SpO2 accuracy under
clinical low perfusion conditions.
79
Table 12. Unpublished clinical trials for GE Healthcare pulse oximeters (continued)
Authors
Device/technology
Study design
Results
Limitations
Raley 2004
[57]
Datex-Ohmeda TruSat
• Population: 10 healthy, non-smoking subjects.
• During stationary conditions, accuracy root mean
square was 1.7.
• Aim: To evaluate the SpO2 accuracy and precision
of the Datex-Ohmeda TruSat pulse oximeter during
motion and non-motion conditions over a wide
range of arterial blood oxygen saturations as
compared to arterial blood co-oximetry.
• Method:
-
Each subject was cannulated with an indwelling
catheter in the left radial artery.
-
The opposite hand was tasked with one of 3
defined motions: mechanically induced tapping,
clinical rubbing motion randomised with the hand
in the prone position or randomised motion with
hand in the supine position.
-
3 pulse oximeters were connected to finger sites
on the motion hand using OxyTip+ sensors. 2
pulse oximeters and the desaturation control
oximeter were attached to the stationary hand on
the cannulated side.
-
Each subject was desaturated to various stable
saturation levels of approximately 4%
decrements from 95% down to 70% and back up
to 100% SaO2 by controlling the fraction of
inspired oxygen.
-
SpO2 values from pulse oximeters were collected
simultaneously to blood being drawn from the
arterial line.
• During motion artefact, accuracy root mean square was
2.5.
• The SpO2 accuracy of the TruSat during stationary and
motion artefact conditions was within specification as
compared to co-oximetry.
80
Table 13. Unpublished clinical trials for Konica Minolta pulse oximeters
Authors
Device/technology
Study design
Results
Limitations
Feiner et al
2002 [45]
Minolta Pulsox-2
• Population: 12 normal subjects.
• Mean bias: 0.17
• Aim: To compare Pulsox-2 with co-oximetry.
• Root mean square: 1.87
• Method:
Feiner et al
2005 [46]
Konica Minolta Pulsox300i
-
Study undertaken by Hypoxia Research
Laboratory at the University of California
Medical Centre in San Francisco.
-
Each subject underwent a standard breathdown protocol of inducing hypoxaemia. Inspired
gas mixtures were controlled to attain hypoxia
plateaus between 70% and 100%.
-
Each plateau was maintained for at least 2
minutes while pulse oximeter readings and
arterial blood samples were acquired
simultaneously.
• Aim: To compare Pulsox-300i with co-oximetry and
determine the effect of different probes.
• Method:
-
Study undertaken by Hypoxia Research
-
-
simultaneously.
• Only a summary of the
report was available.
• Four probes were tested with the Pulsox-300i and the
mean bias ranges from -0.35 to 0.23.
• The root mean square for the four probes ranged from
0.95 to 1.73.
• Only a summary of the
report was available.
• Combined the mean bias was -0.02 and the root mean
squared was 1.28.
81
Table 14. Meeting abstract of clinical trials for Masimo pulse oximeters
Authors
Device/technology
Study design
Results
Limitations
Goldstein et
al 2003 [47]
Masimo Radical
• Population: 19 neonates.
• Philips oximeter zeroed out and gave no signal for more
than 10 times as many instances and more than 25
times as long as Masimo.
• Aim: To ascertain the clinical utility of the Masimo
Radical and Philips Viridia.
• Method:
- Oximeter probes were placed according to
manufacturer specifications. ECG HR was
recorded to compare with oximeter’s PR.
- Criteria of for evaluation were: false desaturation
(reading of <85% was not corroborated by other
oximeter or physical findings); dropout (oximeter
gave no SpO2 and PR readings); changes in HR
(PR differed by more than 25 bpm from ECG
HR).
Mottram et
al 2005 [48]
Masimo Rad-57
• Population: 31 subjects who required blood gas
analysis for clinical evaluation.
• Aim: To compare SpO2 and SpCO measured by
Rad-57 with SaO2 and SaCO measured by a cooximeter.
• Method:
- Measurement from Rad-57 were recorded when
blood sample was being drawn.
- Blood was analysed with 15 minutes using
Radiometer ABL 725 analyser according to
standard laboratory practice.
• There were significant differences in the false
desaturations of the two device, as well as significant
differenced between the devices in erroneous PR as
compare to ECG HR. In false desaturations and
erroneous PR, Philips exceeded Masimo by greater
than twofold instances and threefold duration.
• SaO2 measured 90.8 ± 5.4% (range 97.5-74.6%)
compared with SpO2 of 93.8 ± 1.8% (range 99-80%).
Paired t-test gave a P-value of <0.001.
• SaCO measured 2.0 ± 1.8% (range 9.3-0.8%) compared
with SpCO of 2.5 ± 2.0% (range 11-1%). Paired t-test
gave a P-value of <0.015.
• Paediatric population.
• Did not consider the
accuracy of the SpO2
measurement of the 2
pulse oximeters, as no
comparison with cooximeter measurements
(gold standard).
• Bias and precision of
SpO2 and SpCO were not
determined.
82
Table 14. Meeting abstract of clinical trials for Masimo pulse oximeters (continued)
Authors
Device/technology
Study design
Results
Limitations
Shah et al
2005 [49]
Masimo SET Radical
• 189 motion tests during normoxia and hypoxia were
performed.
• Not-peer reviewed.
• Missed events were counted for the 54 events during
hypoxia. Masimo had the fewest missed events (1/54)
and Nonin had the greatest (22/54).
(gold standard).
• Aim: to compare Masimo SET Radical, Philips CMS
and Nonin 9700 under conditions of low perfusion
and motion in hypoxic and normoxic states.
• Method:
- Low peripheral perfusion was induced by
lowering the room temperature to 16-18°C.
- Test hand was placed on motor-driven random
motion table to induce rubbing and tapping
motions.
- Masimo ear sensor was used as a control during
hypoxia studies.
- Hypoxia was induced using a re-breathing circuit.
- Data was recorded during normoxic and hypoxic
conditions. False alarms were noted when SpO2
dropped below 90% during normoxic conditions.
A missed event was defined as the inability of the
monitor to recover after desaturation, by the time
the control monitor reached 100%.
• False alarms were counted for 135 room air motions.
Masimo has the fewest false alarms (1/135) and Nonin
had the greatest (18/135).
• The sensitivity and specificity of the Masimo were
significantly higher than the sensitivities and specificities
of the Philips and Nonin.
83
Authors
Device/technology
Study design
Results
Limitations
Shah and
Estanol
2006 [53]
Masimo Radical
• 160 motion tests were performed: 120 on room air and
40 during desaturation. Missed events were counted for
the desaturation episodes and false alarms were
counted for the 120 room air motions.
• Aim: To determine which of the Masimo Radical,
Nellcor N-600 and Datex-Ohmeda TruSat were the
most reliable and accurate during difficult patient
conditions.
• Method:
- Each oximeter was connected the test hand and
control hand of each subject.
- Motion was random self generated and machine
generated with the test hand attached to a
motion table.
- A re-breathing circuit was used to induce
desaturation to approximately 75%.
- A missed event was defined as the inability of the
pulse oximeter to detect desaturation and/or
recover from a desaturation by the time the
control reached 100%.
- A false alarm was recorded during normoxic
phase and defines as a SpO2 ≤90% during
motion.
• Masimo Radical missed the fewest events and had the
lowest number of false alarms.
• Datex-Ohmeda TruSat missed the most events.
• Nellcor N-600 had the most false alarms.
(gold standard).
84
Authors
Device/technology
Study design
Results
Limitations
Shah and
Estanol
2006 [52]
Masimo Radical
• Masimo significantly outperformed the Nellcor and
Datex-Ohmeda pulse oximeters.
• Nellcor spent the greatest percentage of time with SpO2
and PR measurements outside 7% and 10% of control.
It also spent the greatest percentage of time with no
signal or zero measurement for SpO2 and PR.
(gold standard).
• Aim: To test the accuracy of the Masimo Radical,
Nellcor N-600 and Datex-Ohmeda TruSat during
motion and induced low perfusion in both normoxia
and hypoxia.
• Method:
- Each oximeter was connected the test hand and
control hand of each subject.
- During separate room air and desaturation
events, motion consisted of random tapping and
random rubbing. Motions were both machine and
subject generated.
- A computer recorded SpO2 and PR data.
Parameters analysed included: percentage of
time SpO2 was within 7% of control; percentage
of time PR was within 10%; percentage of time
SpO2 and/or PR were zero or no signal;
performance index.
Cox 2007
[54]
Masimo SET Radical with
LNOP Blue sensor
• Population: 12 paediatric patients.
• Aim: To compare the accuracy of the Masimo
Radical with LNOP Blue sensor and Nellcor N-600
with Max-I LoSat sensor with a traditional pulse
oximeter on cyanotic cardiac lesion patients in ICU.
• Method:
- Arterial blood gases were obtained as clinically
needed and compared with the pulse oximetry
readings from the 3 devices.
- SpO2 and SaO2 were compared using linear
regression and accuracy in terms of RMS (ARMS).
• 60 blood gas measurements were compared with SpO2
in patient population. SaO2 measured 72.3 ± 7.3%
(range 85-56.1%).
• For the Masimo, SpO2 measured 70.5 ± 7.5% (range
87-52%). Bias was -1.91, precision 3.50 and ARMS 3.97.
• For the Nellcor, SpO2 measured 75.9 ± 5.6% (range 8961%). Bias was 3.81, precision 5.26 and ARMS 6.49.
• For the traditional pulse oximeter, SpO2 measured 75.2
± 6.4% (range 91-57%). Bias was 1.86, precision 6.24
and ARMS 6.51.
85
Table 15. Unpublished clinical trials for Mediaid pulse oximeters
Authors
Device/technology
Study design
Results
Limitations
Mediaid
2006 [44]
Mediaid technology
• Graph showing the correlation between the SpO2
measurements made by Mediaid technology with SaO2
measurements made by co-oximetry.
• Aim: To compare Mediaid technology with cooximetry.
• Method:
-
Study undertaken by Anesthesia Research
-
-
simultaneously.
• ARMS values not quoted.
Table 16. Posters of clinical trials for Nellcor pulse oximeters
Authors
Device/technology
Study design
Results
Limitations
Ahrens and
Ott 2006
[50]
Nellcor OxiMax N-600
• Population: 100 critically ill patients in ICU.
• There were no statistical differences between the three
systems in the display of PR.
• Aim: To compare Nellcor OxiMax N-600, Masimo
SET Radical and Philips Intellivue Fast pulse
oximeters with co-oximetry.
• Method:
-
SpO2 and HR were compared for each pulse
oximeter with co-oximetry and ECG HR.
-
Follow sensor placement SpO2, PR and ECG
HR were recorded under stable conditions prior
and following arterial blood gas sampling for
clinical care.
• Bias and precision of Nellor measurements were 0.18%
and 2.25%, respectively.
• Bias and precision of Masimo measurements were
0.31% and 1.98%, respectively.
• Bias and precision of Philips measurements were 0.19%
and 2.58%, respectively.
86
Table 16. Posters of clinical trials for Nellcor pulse oximeters (continued)
Authors
Device/technology
Study design
Results
Limitations
Sautel 2006
[51]
• Population: 10 infants.
• SpO2 mean difference and standard error between the
Masimo and Nellcor was 1.46 ± 0.36, which was
statistically significant.
• Aim: To compare the ability of Nellcor OxiMax N600 and Masimo SET Radical used on NICU
patients to post oxygen saturation and PR data
during various conditions commonly encountered in
the NICU environment.
• Method:
Lussky et al
2007 [55]
-
Oximeters were attached to either foot of each
patient. Measurements were recorded for 4
hours with 3 hour sensor rotation.
-
Data was capture via laptop.
-
Dropout rate was defined as the percentage of
total monitoring time where data was not
detected by pulse oximeter.
• Population: 17 newborn infants with umbilical
arterial line.
• PR mean difference and standard error between
Masimo and Nellcor was -1.21 ± 0.74, which was not
statistically significant.
• Overall the Masimo dropout rate was statistically higher
than the Nellcor.
• SpO2 and PR for both devices were less than 1% of the
recorded time, but Masimo SpO2 was 1.94 times more
likely to drop out than Nellcor and Masimo PR was 1.55
times more likely to drop out than Nellcor.
• There were statistically significant correlations found
between SpO2 measured by both oximeters and SaO2,
but no significant difference between the performance of
oximeters.
• Aim: To establish accuracy and dropout rate for
Nellcor OxiMax N-600 and Masimo SET Radical
pulse oximeters for use in NICU.
• Bias and precision of Nellcor was 1.76 ± 3.25%.
• Method:
• Bias and precision of Masimo was 1.39 ± 2.81%.
-
SpO2 was continuously measured by oximeters
for 4 hours, with sensor rotation at 2 hours.
-
SpO2 and PR data were exported to a laptop.
-
SaO2 was measured using co-oximetry.
• Nellcor has a statistically lower dropout rate than the
Masimo.
• Paediatric patients.
(gold standard).
• Paediatric patients.
87
Table 17. Unpublished clinical trials for Nonin pulse oximeters
Authors
Device/technology
Study design
Results
Limitations
Severinghaus
2005 [39]
Nonin ONYX II
• Total number of samples was 253, giving an overall
ARMS of 1.31 %SpO2, for an SaO2 range of 70-100%.
• Aim: To compare the accuracy of Nonin PureSAT
ONYX II with gold standard of co-oximetry.
• Method:
-
Each subject was placed in a semi-supine
position and allowed to breathe through a
mouthpiece while nose was block with a nose
clip.
-
Hypoxia was induced by subjects breathing
mixtures of nitrogen, room air and carbon
dioxide, with each desaturation level held at a
stable plateau.
-
Each subject was exposed to 2 runs of 5
-
At each level SpO2 and SaO2 were measured.
-
Measurement of accuracy was presented as
ARMS and calculated according to ISO
9919:2005.
• The performance of the ONYX II was shown to be
consistent over the oxygen saturation range.
88
Table 17. Unpublished clinical trials for Nonin pulse oximeters (continued)
Authors
Device/technology
Study design
Results
Limitations
Nonin 2006
[40]
Nonin OEM III module
(Nonin PureSAT)
• ARMS for SaO2 ≥60% was 1.70.
• Aim: To compare the accuracy of Nonin PureSAT
signal processing technology against the gold
standard of co-oximetry during patient motion.
• ARMS for SaO2 ≥70% was 1.68.
• Method:
• SpO2 measured by Nonin system correlated well with
SaO2 across changing levels.
-
Each subject was placed in a semi-supine
position and allowed to breathe through a
mouthpiece while nose was block with a nose
clip.
-
Hypoxia was induced by subjects breathing
mixtures of nitrogen, room air and carbon
dioxide, with each desaturation level held at a
stable plateau.
-
Each subject was exposed to 2 runs of 5
-
At each level SpO2 and SaO2 were measured.
-
Rubbing and tapping motions were created by
means of an industry-wide accepted protocol for
generating motion.
-
Measurement of accuracy was presented as
ARMS and calculated according to ISO
9919:2005.
• At saturation levels below 80%, Nonin performed
consistently.
• No misleading data was observed.
89
Authors
Device/technology
Study design
Nonin 2007
[41]
Nonin Onyx 9500
University of
California San
Francisco
2008 [42]
Nonin Avant 9700
Results
Limitations
• ARMS for Nonin and SPO was 1.4 and 4.2, respectively.
• Aim: To assess the accuracy of the Nonin Onyx
9500 and SPO Medical PulseOx 5500 compared to
arterial blood saturation levels in normal subjects.
• The accuracy of the Nonin was consistent throughout
the range of 70% to 100% with no signal deterioration at
the lower ranges of oxygenation.
• Method:
• The SPO oximeter’s accuracy decreased at values less
than 90%.
-
Hypoxia was induced to 5 levels of oxygen
saturation between 70% and 100% by having
subjects breathe a mixture of inspired gas.
-
Each level was held at each plateau for
approximately 2 minutes whilst pulse oximeter
readings and arterial blood gas samples were
taken simultaneously.
-
25 samples were obtained for each subject.
-
Accuracy was defined as the closeness of
agreement between a test result and an
acceptable reference value.
• Aim: To compare the accuracy of Nonin Avant
9700, Masimo Radical and Nellcor OxiMax N-595
and determine the effect of dark skin on accuracy.
• Method:
-
Each subject underwent a standard breathdown protocol to achieve arterial oxygen
saturation between 70% and 100%.
-
Accuracy was reported as ARMS.
• At 70% and 80% oxygen saturation, the mean bias in
dark skin pigmentation was minimal for Nonin’s oximeter
at -0.6% ± 1.4, compared to a mean bias of 2.1% ± 2.9
for Masimo and 2.0% ± 2.1 for Nellcor.
• ARMS for Nonin, Masimo and Nellcor was 1.3, 3.6 and
2.9, respectively.
• Nonin maintained acceptable variability in the most
challenging environment of dark skin pigmentation and
SaO2 less than 80%.
90
Authors
Device/technology
Study design
Results
Limitations
University of
California San
Francisco
2009 [43]
Nonin WristOx 3100
• ARMS for Nonin, SPO and Minolta was 1.7, 3.4 and 3.1,
respectively.
• Aim: To compare the accuracy of Nonin WristOx
3100 with SPO 7500 and Minolta PULSOX 300i
pulse oximeter.
• Method:
-
Each subject underwent a standard breathdown protocol of induced hypoxaemia. Inspired
-
minutes while pulse oximetry readings and
simultaneously.
-
25 samples were obtained for each subject.
• The accuracy of the Nonin was significantly better than
the other wrist pulse oximeters tested.
• The accuracy of the Nonin was consistent throughout
the range of oxygenation. Conversely, the accuracy of
the other devices decreased at values less than 90%.
Author and report information
Market review:
Pulse oximeters in primary and
prehospital care
E S Colechin, D R Bousfield,
C A Reay and A J Sims
Regional Medical Physics Department
Freeman Hospital
High Heaton
Newcastle upon Tyne
NE7 7DN
Tel: 0191 213 7787
Fax: 0191 213 0290
Email: [email protected]
Website: www.rmp.org.uk
About CEP
The Centre for Evidence-based
Purchasing (CEP) is part of the NHS
Purchasing and Supply Agency. We
underpin purchasing decisions by
providing objective evidence to support
the uptake of useful, safe and
innovative products and related
procedures in health and social care.
CEP publications since 2002 are
available from our website.
www.dh.gov.uk/cep
© Crown Copyright 2010
91

Market review: Pulse oximeters in primary and prehospital care

Transcription

Similar documents

ErgoPulse 25PTX Impulse Nutrunner

Onyx® II Model 9550 Finger Pulse Oximeter

Your Health: How to Measure Your Pulse

Nonin Finger Pulse Ox

ES-80L - Mega Elektrik Shop

Pulse Cereal Partnership

Schnitzler - Swissphotonics

Educating the Educator: Use of Pulse Oximetry in Athletic Training

DICROTIC NOTCH IN FEMALES

Initial Operation of a Pulse-Burst Laser System for High